In order to meaningfully compare costs and benefits of lockdown, it is necessary to have some assumptions about the value of human life. There are different approaches that can be taken to this, such as using the (quality-adjusted) life year thresholds used by public heath systems to ‘price’ a year of healthy human life. Of course, it is popular and politically expedient to ignore the price of life issue, indeed some like to argue that human life is of infinite value. This is nonsense. Human life has a finite value, else we would have no basis to determine whether it made sense to approve treatments based on their cost and their medical merit.
Modelling life expectancy in terms of GDP
The approach I will take is slightly different here. I will attempt to model the relationship between GDP and life expectancy using a simple regression model for the EU-28 countries. The intuition is that higher GDP enables a prolongation of human life, through better diet, investment in workplace health and safety and investment in curative and preventative medicine. Over the relevant part of the curve, we should see a clear relationship emerge, which will allow us to understand what the cost in terms of life is, of reducing GDP. This can then be compared with the estimated benefits – also in terms of human life – of implementing a lockdown.
I use the data from Eurostat for 2018 GDP per capita in euros, (GDP), for life expectancy at age 1 (LE) to construct a simple regression model specified as follows
LE = a*GDP+b*GDP^2
The spreadsheet with the data is here. I exclude Luxembourg and Ireland because of well-known problems with the measurement of GDP. While other control variables are not included, in fact the institutional nature of the EU suggests that many other relevant factors (such as the design of the public health system, workplace health and safety rules, and other cultural factors) enjoy a certain degree of harmonisation.
The results of the regression are as below:
And show that both variables are statistically significant (high t Stat values), and that quite alot of the variation (around 60%) in LE can be explained by GDP/capita alone (high ‘Adjusted R Square’).
The coefficients in the last three rows imply:
(a) that in the absence of any GDP, life expectancy would be at 70 (intercept). Although this seems unreasonably high, suggesting more non-linearity in the relationship as income falls to the zero-threshold, it is not a big deal, however, as we are only concerned with the relationship over the relevant space, i.e. in terms of the GDP/capita of developed, Western countries.
(b) there is a positive relationship between GDP and Life Expectancy; namely that a one Euro increase in GDP/capita results in an increase in life expectancy of 0.0007 years. This is what we would expect to see.
(c) there is a negative relationship between GDP squared and Life Expectancy; namely that as the square of GDP rises by one euro, LE falls by 0.000000001 years. This is also what we would expect – the gains to life expectancy are limited, and at very high income levels, there may even be a reverse effect, due to stress from work, pollution or other lifestyle effects.
The graph below shows the scatterplot of actual values and the estimated line of the curve from the regression:
In summary, on average, a 5% reduction in GDP will result in a loss of an estimated 0.5 years, or 6 months, in the average life expectancy of people in Europe. This will happen because of many reasons, such as reduced spending on curative health, increased crime, more suicides, greater occupational risks and reduced spending in preventative health and safety.
Whether lockdowns make sense or not is an important question, and it is hard to answer definitively. Hard, for a number of reasons. First, as is often said, we do not have all the data at our disposal. On the epidemiological side, we are only just beginning to understand how the virus has been spreading and whether or to what extent lockdowns make a difference.
Quantifying the economic costs of lockdown
On the economic side, there is a large degree of uncertainty. While clearly the economic costs are large, there are inherent difficulties in untangling the costs of lockdown per se from the costs of the general COVID hysteria which would ensue, even in the absence of any government measures. And even then, calculating the costs of lockdown itself is tough. As economists are at pains to stress, this episode is without precedent in modern times, and given the extent to which Western economies have developed over the past fifty years, the older precedents such as WWII or the Great Depression, lose relevance due to the differences in how modern economies operate and the level of prosperity we have come to take as a baseline.
So we have to really try to go back to first principles. What have lockdowns done? As I see it, they have done two things to the economy. First, they have stopped production. Quite simply, governments have closed stores, factories, restaurants, bars, schools, etc., meaning there are fewer goods and services being produced by the economy. How much of the economy is now closed depends on the nature of the lockdown (France, Italy, Spain, Belgium, Germany, New York City, Michigan, Ireland…, are extreme examples; many other US states and the Netherlands less extreme, Sweden the least extreme), but easy estimates put the number at close to a third of the economy.
In some cases, production will have shifted from the formal to the informal economy: restaurants are no longer preparing food, but people are cooking more at home; teachers are no longer caring for and minding children, but parents are doing so at home, etc. However, in many cases, the production is simply being lost. Again, how much is the result of the lockdown and how much of the generalised COVID panic which would have occured anyway? It’s difficult to say, but to put the loss at 15% of the economy does not seem unreasonable, especially when we consider that the two are interlinked: people panic in part because governments lockdown.
The second effect is the productive efficiency loss. We can take teleworking to be less efficient that working in offices, as otherwise the economy would have phased to telework years ago. (After all, the technology is not particularly new and companies wouldn’t willingly fork out for prime office space if they could get the job done as efficiently with no rental costs.) Home schooling is also a large efficiency loss, related to scale diseconomies (a class of 24 formally taught by one teacher, is now being taught by 24 parents) and the fact that parents are not trained teachers. Supply chain disruptions mean that work stoppages occur even in otherwise unaffected sectors: E.g. construction companies cannot pop down to the local supply shop to get an extra bucket of 3″ screws). Quantifying these efficiency losses is tricky, especially because they may be partially offset by some efficiency gains (fewer useless meetings for people to sit through).
There is also, I think, a non-linearity here with regard to time. If we think of businesses as having both an operational and a strategic component, then the former can probably cope better with remote working for some time, and in the short term, a company can get by without a new strategy. But in the long term, the lack of face-to-face contact will make strategic decisions like restructuring, brainstorming, hiring and firing…, much more difficult. This means fewer new ideas, less innovation, less efficiency gain.
Finally, there is the economic impact of social costs. The harm to people’s mental health from isolation, the loss of emotional development for, in particular, vulnerable children. The scarring effects of unemployment and disattachment from the labour market for those who have lost their jobs. While these are all first and foremost social costs, they will also impact the economy through reduced output, higher crime and wasted human potential. Again, these are non-linear with regard to time: the longer the lockdowns go on, the more costly each added day of lockdown will be.
Quantifying all these effects is not easy. If we take as a baseline four months of effective stoppage, and assume this halts 15% of output, then countries in full lockdown can expect at least a 5% one-off loss to annual GDP from the first effect. For the other two, the effects are likely to be more drawn out, but less acute. Using the four-month baseline again, the efficiency cost of teleworking could be estimated using the rental value of the office space currently being leased. In the EU, there is an estimated 650 million square meters. At an annual rent of €400 per square metre per year, this makes about 90 billion in lost office for the 4 month period, or about 0.5% of GDP. Of course, this is a lower bound; it’s reasonable to assume, given the transition costs associated with the shift, that this number is closer to a full percentage point in GDP.
Next is lost productivity for parents. Here we can assume that working parents of dependent children suffer, as a result of the caring responsibilities. Data is available from Eurostat on employment rates of parents by number of children, the distribution of population by number of children. Combined with data on industrial annual earnings for the age group 30-39 (a proxy), we can estimate the lost productivity using wage data and assuming single parent households lose 50% of productivity, two parent households with one or two children lose 25% of productivity and households with three children lose 10% of productivity. (Households with more than two adults are assumed not to lose productivity). This amounts to another €90 billion for the 4 month period, another 0.5% of GDP.
Other costs arise from the stoppage of routine medical activity and the mental health toll. Here indeed, disentangling lockdown from baseline costs is tricky. Undoubtably, any alternative to lockdown would carry with it a certain mental toll, as well as an increased (albeit likely short term) demand on health services. Similarly, other social costs such as the damage to disadvantaged children cannot be readily quantified, but are nevertheless important qualitative considerations.
In summary, the upfront costs of a four month lockdown can be roughly estimated to be between 15 to 20% of GDP, with scarring effects causing a permenant reduction of GDP that is far more difficult to quantify – not least because it will depend on the policy response. Given these impacts on human capital, on educational outcomes for children and on mental and physical health, it seems difficult to imagine a swift V-shaped recovery. In present value terms, a permenant drop in the order of 5% does not seem like an unreasonable working assumption.
It seems that the public debate is divisive as never before. Left versus right. Social Justice Warriors (whatever they are) versus the alt-right (whatever that is). Admittedly, this impression may simply be a result of our unhealthy addiction to social media, which through its Algorithms of Hate and its cloak of anonymity tends to radicalise latent tendencies. It drives us into tunnels with those who share our vision, while at the same time shielding us from the consequences of our push-button outrage.
Nonetheless, I would contend that the fabric of culture is indeed in the process of tearing. The seams of our civilisation – things like basic human dignity, kindness, respect, humanity and tradition – now seem incapable of holding us together against the pressure of our outrage. The space for reasoned debate seems to shrink with every passing tweet.
A change is as good as the rest
The trouble is that the two sides are not even consistent in their own viewpoints. In the United States, the Democrats – formally defenders of the economically disadvantaged – have somehow come to represent an uneasy alliance between middle- and upperclass privileged ‘Coastals’ and traditional urban ethnic minorities, while the Republicans draw support from the equally awkward bedfellows of the superrich and the underclass of rusting, undereducated, mostly white ‘Heartlanders’. Neither side consistently upholds the values of the left (i.e. a larger state, more redistribution, protection of workers rights) or the right (more free market, less regulation, lower taxes and a smaller state). Members of a group which formally advocated liberal values such as free speech now rally behind brutal authoritarian slogans like “punch a Nazi”, failing to appreciate the irony of their own intolerance.
In Europe, mainstream social democrats increasingly resemble historical anachronisms, while the (Christian democrat) centre-right, desperate to hold on to a collapsing middle, is being overtaken by populists. The CDU in Germany has through the open borders refugee politics of Angela Merkel betrayed many of its own base, while the centre-left’s failure to check the consequences of rising inequality and globalisation constitutes an equal betrayal in the eyes of the old working class faithful. In Britain, the single-issue of Brexit has torn apart both the Conservatives and Labour. In its place, it has forged an unlikely anti-EU alliance between post-industrial working class northerners and well-off village conservationists in the affluent Home Counties which surround the (anti-Brexit) London metropolis. Only yesterday, Italy joined in the fun by voting in populists and extremists and booting out the moderate centre-left. But nothing illustrates the collapse of the old left-right dichotomy as forcefully as the 2017 French presidential election, which saw Emmanuel Macron’s virgin movement sweep aside both the PS and the Gaullists in his ascent to the Elysée throne, although the real victor was Madame l’Absention, followed closely by Madame Le Pen.
New times call for new political structures
The problem is, the model of the old left-right divide has always been missing something important. It was only ever the particular circumstances created by industrialisation which allowed for this one-dimensional approach to work as a good approximation. The left-right split also had the added advantage of convenience, in that the institutions of representative democracy work best when competition for political power is limited to a few, slightly differentiated brands. Being able to position two, three or four parties on a single spectrum reduced the task of political choice to something the masses could participate in without much active engagement or intelligence. The accidental balance created by industrialisation permitted us to overlook the fact that the model we were using to think about society was fundamentally flawed. As the world evolves, that balance is increasingly disturbed. We need to develop a new framework. As the world becomes more complex, so too much our political framework.
But how to adapt the model? Ideas abound; a popular version adds another ‘ecology’ dimension to the old left-right divide. However, in practice, environmentalism is less a coherent ideology and more a loosely grouped set of specific policy challenges. Once you break it down, the solutions to environmental problems can be tackled by taking a position on the conventional left-right spectrum. For instance, if you’re a left-winger, pollution taxes which internalise the externality seem like a good idea; if you’re a right-winger, you might like tradable pollution permits which achieve the same result. Nor does the environment feature particularly strongly in the ideological divide that is tearing the Western world apart at present.
The Social Triangle
I propose instead a model which takes into account the dimension I feel has been missing. As the diagramme illustrates, we can see society as consisting of three dimensions, the two which are conventionally understood in existing political analysis (the State and the market) and a third, that of community. Community can be understood as the voluntary organisation of people in groups without a specific transactional motivation (which is the market) and without the element of compulsion/threat of violence (which is the state). I argue that the absence of this dimension has, up to now, remained unobserved because we have in essence taken the role of community for granted. It is only as communities collapse – i.e. as society begins to drift ‘downwards’ towards the bottom side of the triangle – that we notice its absence. It is marked by a tendency towards the unholy alliance between the state and the market, in the form of corporatism.
When looked at this way, society can be properly understood to be positioned somewhere within the above triangle. Societies that are more communistic can be placed within the red area of the triangle; those that are more anarchistic within the black area; and those that are more corporatist within the blue area. An ideally functioning society, like a three-legged table, balances all three dimensions and ends up somewhere in the green centre. Of course, in today’s world, we are observing a collapse of community – a massive drop-off in religious subscriptions, fewer bowling clubs, the death of the Boy Scouts, etc. and so we have drifted downwards, further towards corporatism. In such a world, the negative effects of the free market become more apparent as, for instance, the lack of community leads to an erosion of business ethics. But equally, a lack of community (for instance in the form of charities) heightens reliance on state-provided forms of welfare, revealing the innate corruption and incompetence of the state’s bureaucracy. Supporters of Donald Trump scream at supporters of Hillary Clinton and vice versa, but in the end, they are both suffering from the same affliction; a lack of community. The rise of populism is not a cry for more or less free market, nor is it a cry for a bigger or smaller role of the state. It is a cry for more community.
Which leads to the next question: What is leading to the erosion of community and how can we reverse this? That is perhaps the subject of another blogpost.
The official website for statistics of the European Union is called Eurostat. If you visit their website, and I highly suggest you do, you will find among the many interesting databases one for the Gender Pay Gap, which measures the difference in average gross hourly earnings between men and women for all European countries (and a few other places too). The message you’re supposed to get from this is that men get paid more than women for doing the same work.
This measure is more than just a statistic, it is a political construct. Despite the fact that it doesn’t really mean what people think it means, the GPG is much quoted in the media, by feminists and even by mainstream politicians. It helps that it has become a sort of catch-phrase, earning its very own capital letters. Because we all know that when a concept becomes capitalised, it Must Be True. And so misread, the Gap sparks outrage, even to the point of politicians passing rather draconian rules to try and level the playing field.
Don’t let reality get in the way of a good story
However, when you start to take a closer look at the Gender Pay Gap in its crude and unadjusted form, for instance by taking into account differences in risk-taking, career choices and hours worked, the difference in pay falls dramatically. And then there’s the fact that the measure tells us nothing about household income distribution. True, men often opt for longer hours (at higher pay) than women because they have wives back home to take care of the kids, but these same men are bound by custom and by the law to share their extra earnings with their families. The value of interhousehold transfers far exceeds the earnings gap, and you can see this by looking at this handy spousal maintenance calculator, which tells us that when it comes to the divorce, a man in the UK earning £40,000 a year will be due to pay his wife, if she earns £30,000 a year, £5,000/year in maintenance. That difference wipes away the benefits he might be getting from the Gap in one fell swoop.
Another way of seeing this is by looking at differences in net disposable household income by gender (which pools household income, and takes into account the value of taxes, and transfers received from the government – in this way we can compare, say, a single female householder to her married male colleague who has two kids to support). This gap is much smaller than the GPG would suggest, and among younger households it is almost zero.
But don’t take my word for the fact that the Gap is meaningless. Consider the view of one group who rarely consider any ideology beyond the Almighty Dollar: Capitalists. Here we see that market researchers are not fooled by the policy bias. They know that women make up 85% of all consumers. Sure, a lot of this is spent on stuff for the family, but even for the pure luxury good market – leisure activities, fancy clothes, dining out, perfumes and chocolate – the gender bias is evident: Men earn the money, but women spend it.
Mind the Gap
Yet that somehow doesn’t slow the narrative, nor dampen the outrage. The feminist policy lobby argues that even when you control for hours worked, career choices and risks taken, a Gender Pay Gap remains (admittedly then much smaller) . And this, they exclaim, is pure discrimination, and must be stomped out by all means necessary.
If you were to suggest the remaining gap can be explained by men – high on testosterone – being simply better at competing in high-value, high-stress work environments, you would be branded a male chauvinist pig. Fearful of this kind of branding, I’ll not dare to suggest such a thing myself.
Check your privilege, and then man the hell up!
Yet weirdly, testosterone is exactly the explanation which is tossed about whenever it comes time to discuss a far more pressing, far clearer gender discrepancy – the gap in mortality between women and men. Here, it is the very risk-taking which leads to higher rates of occupational accidents, which in turn kill off men at a faster rate than women. In other words, if a man takes risks and dies for it, he has only his toxic masculinity to blame. If he takes risks and gets paid more for it: INJUSTICE!
The Gender Mortality Gap is, as far as I can tell, a phrase coined by me. Normally I’m proud of being able to claim credit for stuff, but in this case I find it alarming that the most obvious, most enduring gender injustice on the planet needs a third-rate occasional blogger like me to invent its catch-phrase. No entry in Eurostat. No Barack Obama waxing lyrical about giving our men back their lives, from behind the Presidential pulpit. Just a guy with a receding hairline and a WordPress site.
Only the men die young
Yet the GMG is real. Though Eurostat doesn’t have a special table for it like for the Gender Pay Gap, but you can still go to their website and calculate the difference in average life expectancy for women and men: For the EU as a whole, the GMG for newborns is 5.4 years, though it varies from 3.3 in the Netherlands to 4.8 in Germany; 6.3 in France; and as high as 10.5 in Lithuania (the US is close to this higher number too). And unlike the figures for hourly pay, which don’t take into account self-employment or black market earnings, statistics on death are among the most reliable we have. There is no trickery here: If you’ve had the misfortune to be born with a penis, you’re probably going to die 5 years younger than your twin sister.
So while the policy world is busy imagining ways to force the free market to pay women more, nobody is asking about what policies are needed to help men live longer. Fortunately there is a clear answer: Spend more money on public health, preventative and curative healthcare for men. The imperative to do so is all the more striking when we consider the fact that men, as the majority taxpayers in countries that publicly fund universal health systems, don’t even get an equal share of the treatments they are shelling out for. The statistics are scant, but the OECD reckons that €1.12 is spent on women’s healthcare for every Euro spent on men’s (this excludes the cost of reproductive treatments such as maternity, pre- and post-natal care).
Economists love to talk about ‘externalities’. ‘Externality’ is a wonderfully complex-sounding word that makes you feel more intelligent just by saying it. It is especially useful when trying to pull the wool over the eyes of a non-economist, as in:
Community activist: “We’re outraged that our organisation has had its budget cut so the government can bail out irresponsible banks!”
Economist: “I understand completely your feelings, but your analysis of the cost of financial sector repair fails to take into account the growth-enhancing effects of the associated positive externalities resulting from the smooth operation of financial markets.”
Community activist: “I… well… uh… what?”
Economist: *smiles imperceptibly and adjusts knot on his silk tie*
Keep your market transaction to yourself, buddy
Yet the actual meaning of the word ‘externality’ is in fact quite simple. Here it is in a nutshell: For any transaction, there is a buyer and a seller. An ‘externality’ can be thought of as the effect of the transaction on someone who is not the buyer and not the seller.
So for example, if John buys a car from Volkswagen, this transaction has effects on both John and Volkswagen (John gets a new Golf in exchange for €30,000; Volkswagen gets €30,000 in exchange for a Golf). Because the transaction was voluntary, we can assume both John and Volkswagen are both strictly better off from having made the trade (otherwise, they probably wouldn’t do it).
But what happens when John tools down the road in his new Golf, kicking carbon emissions and particulates into the air, taking up public space and potentially running over grannies and cats? In that case, there is a cost born by someone who was not party to that transaction, arising from the transaction (pollution, traffic congestion, increased risk of an accident). This cost is a ‘negative externality’. There are many examples of negative externalities, but pollution is perhaps the most common one. In general, economists accept that the government should sometimes intervene in markets in order to correct for these negative externalities (through regulation or taxation, for instance).
There’s no such thing as a free ride…or is there?
We can also think of ‘positive externalities’, i.e. when the operation of a market has a positive effect on someone outside it. The pleasant smell of freshly baked bread on a street outside a bakery can bring joy to passers-by, even if they do not actually enter and pay for the bread. And if a person with a contagious disease pays to have himself treated privately, this is a benefit to all the people he has protected from potential infection, even if he was only acting selfishly.
But don’t let the name deceive you: positive externalities are not always a good thing. Sometimes they can stop markets from operating effectively. For example, if I invent a brilliant new machine and try to sell it, the ‘idea’ can simply be copied by someone who does not actually pay for my machine, leaving me with only a small reward for all the midnight oil I burned while getting to my eureka moment. Without some kind of protection, this risk might prevent me from bothering to invent the machine in the first place. This is why, just as in the case of negative externalities, the risk which positive externalities pose to production is a justification for the government to intervene, by granting patents and other forms of intellectual property rights.
Compound externalities – where markets depend on failure in order to succeed
There is a certain class of externality which I find quite interesting and which, to my knowledge, has not been written about by economists yet. It is the ‘compound externality’, which can be defined as the effect on someone outside a market arising from the operation of a market which only has value to the buyer and seller because of this negative effect. This sounds confusing, but it’s quite simple when we break it down: As before, we have a transaction, so we have a buyer and a seller. In addition, there is an effect from the transaction on a third party AND – here’s the catch – the only reason the transaction is valuable to the buyer / seller is because of this effect on the third party.
The most obvious example of a compound externality is noisy alarms. In this case there is a negative cost imposed on you when your neighbour’s house alarm goes off at 3 in the morning, even though you didn’t sell him the alarm and you sure didn’t buy it. But here’s the catch: the only reason he wanted the alarm is so that it would annoy and wake you, his neighbour, up, so that you would then look out the window at his house and – in this way – spot the burglar breaking in through the window (hence deterring the burglar!). If it didn’t create the noise pollution, the alarm would have no value.
Another example is the ‘ugliest façade’ project. Imagine you live on a historic square full of old houses with lovely façades. Your obnoxious neighbour from the above example – not content with his cacophonous alarm – has torn down his house and is building anew. He has an incentive to build the façade as ugly as he possibly can. Why? Because by doing so, he destroys the perfect appearance of the square as viewed from your house and from every other building… every building that is, except his own. All of the sudden he possesses the only piece of real estate with an unsullied view of the historic old square!
Quick, is there a positive compound externality in the house?
What about positive ‘compound externalities’? Are there any examples of markets which, in order to operate, depend on a positive effect occurring on someone outside the market in order for the market to exist? The only one I can think of is the market for first-aid training. Here, the only value to you in paying for a first-aid training course is that, should the occasion arise, you will be able to apply the Heimlich Manoeuvre to dislodge a chicken dumpling from the oesophagus of a third party (while attending a party…).
I’ll stop now because I can’t think of any other examples. Perhaps someone else can?
If you know me or have visited this blog before, you’ll know that my book, The Hydra, is about overpopulation. In it, a scientist decides the world is so full of humans, that he must save the planet by engineering and releasing an infertility virus. I won’t give away too much of the plot, but suffice it to say as a novel it doesn’t really do much hard number crunching. It begs -but perhaps never really comprehensively answers- the crucial question: Is the world so overpopulated that we’ll destroy the planet unless we change our policy direction?
Indeed, when you discuss the issue with most people, you get lots of uninformed opinions, which range from “I think we’re all doomed, unless there’s some major war or something” to “There are definitely not too many people in the world. It’s just a question of sharing out the world’s resources fairly and investing in technology instead of war” I always find it astounding just how convinced both sides can be of their opinions, without the faintest notion of what the hard numbers are saying.
So let’s see if we can do any better. First stop, the databank of the World Bank where, after some data cleaning, we can come up with a list of useful data for the world’s countries, grouped into categories depending on how rich they are. Basically we’re looking at four things: population, Greenhouse Gas (GHG) emissions, GDP and birth/death rates, from 1960 to 2015.
What can we see from the numbers?
The first thing to look at is Greenhouse Gas (GHG) emissions. By 2012 we humans had pumped about 567 gigatonnes of CO2 equivalent (GtCO2) into the atmosphere since the dawn of industrialisation (1870). In 2012 alone we added 52 GtCO2 to this stock, up from an annual total of 27 in 1970. While the data are jumpy, on average, the amount of carbon we release in the atmosphere annually is growing by about 1.3% a year. Even if this pace of annual emissions growth were to fall from 1.3% to 0% a year, that would still mean we would be adding the 2012 payload of 52 GtCO2 into the atmosphere every year. If that level of emissions were to continue until 2050, that would result in an atmosphere laden with 2,572 GtCO2 released by humans. That number is so big, it is literally off the charts, as far as the climate scientists are concerned. To illustrate, I’ve made a simplified version of “the chart”, i.e. the UN’s reckoning of how cumulative emissions will raise temperatures. You can find the full chart on page 54 of this document.
As you can see, they don’t even consider a scenario in which we keep emitting the level of GHG which we emitted in 2012. What this is telling us is pretty clear: if we continue with business as usual, we’re going to miss the current climate targets by more than a factor of 2, resulting in massive, truly massive, changes to our climate which may well spell disaster for the planet and for us all. Now of course, nobody believes business as usual is an option, which is why we had Kyoto and then Paris and soon Marrakesh.
Linking emissions to income
Money makes the world go round. And it also determines how much GHG we put into the atmosphere. Or more precisely, the things people like to spend money on: heat, bigger houses, clothes, high-protein food, transport. The precise link between income and emissions depends on where a person is on the income scale: For very rich countries, there is already evidence of ‘decoupling’, i.e. as rich people get richer, emissions don’t increase, they actually go down. But because rich countries only account for 15% of the world’s population, that doesn’t really matter. What matters are the middle income countries, places like China and Brazil, who make up 35% of the world’s population. The 2.6 billion people living in these countries have been getting richer since 1990, and whenever they’ve got their extra cash, they’ve burned it and pumped it into the atmosphere. Here’s the chart that shows it:
As these countries get to the sort of income levels the rich world has already achieved (and they are well on their way) there is every reason to assume that they too will ‘decouple’ emissions from growth, but for now, they are still hungering for more of the things that make the atmosphere hot: steak, cars, swimming pools and city breaks. And this is set to go on into the foreseeable future.
The real problem, though, is the next wave of countries, the so-called “Lower Middle Income” countries like India, which as a group are home to even more homo sapiens (2.8 billion or 40% of the world’s population). If these countries grow in the same way as China and Brazil have done, it will mean even more pressure to emit.
The power of econometrics can help us to estimate this relationship, which turns out to be very well approximated by the equation [kilograms of emissions / per person] = 1090 + (0.7093)*[GDP/person] – (0.0000047025)*[GDP/person]^2 – (0.00000000010531380)*[GDP/person]^3. If you want the nerdy details of where I got this, click here, but for everyone else I’ll just summarise what this means: If you have zero income, you will still emit about 1,000 kg of CO2 equivalent into the atmosphere every year. Emissions go up at about a rate of 700 grams a year for every dollar of extra income you get, but this slows down as your income approaches $35,000 a year. After that, extra income leads to lower carbon emissions per year. Here’s what the graph looks like:
So what about population?
In 2015, there were 7.3 billion humans on Earth, more than ever before. This population increases about 1.2% a year, and while the rate of increase has been slowing since the late 1960s, it hasn’t been slowing by very much. If the pace of population increase we have observed since 1969 were to continue (i.e. let’s assume it continues slowing a bit every year, like it has been doing) there would be 9.75 billion of us by 2050. This, by the way, is the EXACT baseline estimate for the UN’s own population projects, but more about the UN’s numbers in a bit.
The main driver in the growth of populations is the crude birth rate, which measures how many children are born per 1,000 people. It turns out there’s a pretty stable relationship between birth rates and per capita GDP. Crude birth rates have been going down pretty much everywhere in the world, and it’s because of money. Basically, the richer a country, the fewer babies they make. In very poor countries, crude birth rates are around 35-40; as a country gets richer, the birth rate falls to just under 10. Figure X illustrates the relationship, which mathematically can be approximated by this formula: b = (45 * minY^a)/Y^a, where b = birth rate, Y is per capita GDP, and a is a “shape parameter” which is somewhere in the range of 0.2622 to 0.36487 Again, for the nerds out there, all the details are here.
The other thing that affects population is the death rate (deaths per 1,000 population). This too is ultimately a function of cash, but the relationship’s a little trickier because of demographic effects. (For example, Germany’s death rate is higher than Zimbabwe’s, not because Mugabe has better health policies than Merkel, but because, when you break it down, old age is the single worst thing for your health, no matter how rich you are. And Germany simply has a lot more old people than Zimbabwe.)
But here again, statistics can come to our aid. We can isolate the effect of the demographics and when we do, we get a pretty similar relationship as with per capita income. This is the equation that tells the story: d = (minY^a)/Y^a * (1/AGE^g), where d = the death rate, minY is a constant equal to 1,011, Y is per capita GDP, AGE is the percentage of the population aged over 65 and a and g are shape parameters equal to 1.0144 and -2.0097 respectively. In other words, the richer a country’s people are, the lower its death rate. The more oldies are in a country’s population, the higher the death rate.
So to recap, as people get richer, they have fewer babies, but they also tend to live longer, and we can use statistics to estimate by how much this is so for every extra dollar of income they get.
Putting it all together
Equipped with the three sets of estimations we have done above, we are ready to put the whole picture together. The first step is to make an assumption about how per capita GDP might evolve in the future. Of course we don’t know, but let’s imagine it continues to grow at the same annual rate it has been growing from 1990 to 2015, for the four classes of countries the World Bank identifies: high income (e.g. the US and Europe), upper middle income (e.g. Russia, China and Brazil), lower middle income (e.g. India and Indonesia) and low income (i.e. mostly sub-saharan Africa). This is what we would get:
The dashed green line illustrates the ‘decoupling’ threshold, i.e. the point beyond which getting richer no longer causes more per capita emissions. As you can see, while the ‘Upper Middle Income’ countries pass this threshold, the ‘Lower Middle Income’ countries – and remember in population terms these are the big guys – won’t even have got there.
Using our equations which we estimated above, let’s now link this assumed GDP/capita path to what we know about birth rates and death rates and see what that gives us for total population:
Now, there’s an awful lot to say about these “GDP driven” estimates of total population on earth. The first thing is that it gives us an estimate of 11.4 billion for the 2050 population, which is a good 1.7 billion more than the UN’s estimates. When you compare the two sets of projections line by line, you see that the differences are in the two “Middle Income” categories. The UN’s estimates seem to assume that these countries population’s will grow more slowly, driven by a faster decrease in birth rates.
The next major difference is that unlike the UN’s demographic projections, these GDP driven projections show no sign of population levelling off any time soon. Indeed, it seems to imply that for the bulk of countries, there’s a good ways to go until death rates overtake birth rates. (I’m willing to put my hands up and say I’m not a demographer, so maybe there’s things I have missed. For one thing, my modelling takes no account of migration trends. I guess I’m kind of assuming that for the world as a whole, net migration is zero. But as people move from poor to rich countries, their birth rates also change, so it’s possible to argue with my numbers).
Now you might be tempted to say: hang on, you just assumed GDP would grow like that. What if growth levels off? Wouldn’t that solve the problem?
Not really. If we change the model so that there is zero growth in per capita GDP from now until 2050 for all four classes of countries, here’s what we get:
Now the 2050 population is projected at a whopping 13 billion! This is because the lower per capita GDP among the ‘middle income’ countries is driving higher crude birth rates.
Finally, let’s bring this all back to total GHG emissions. If we plug these population numbers into our estimate for per capita GHG emissions by income level, we should be in a good position to tally up the total GHG emissions that this implies, under the two scenarios (no growth and growth at the average).
As you can see, with zero economic growth, the level of emissions is lower, but still really high. (The reason why it stays so high is because, at zero growth, although lower middle income countries like India are not pumping more GHG into the atmosphere, the rich and upper middle income countries still are. Furthermore, while the per capita emissions of the poor stays low, their numbers are increasing at a faster rate.)
For both graphs, the red line indicates the UN’s uppermost threshold (2,310 Gt CO2) for their most extreme emissions scenario. Therefore, the current projections put us on a path of GHG emissions that would mean temperature increases to 2100 of more than 4 degrees Celsius. Once again, well off the charts!
Can technology save our bacon?
It’s entirely possible that sometime next year, “they” will discover cold fusion, a carbon-less, virtually free and infinitely renewable energy source that will allow us to rapidly decarbonise and merrily turn the planet into some kind of Coruscant. I have no clue whether this is a realistic prospect.
But sadly, neither do the people who seem to be depending on it as a solution. And my instincts tell me it is very bad policy to rely on a solution not yet invented in order to solve a problem so grave that it threatens our species’ very existence.
Maybe the way to save our bacon is simply to stop eating it? Lowering our consumption of meat and other carbon intensive goods will surely help. Yet when we look at the scale of the challenge as outlined above, it is clear to me this can only be a part of the solution.
Given that the underlying problem is that there are a lot of people in the world, birth rates are higher than death rates, and most of the world’s poor are getting richer, it seems to me that – absent Cold Fusion – there are really only two other choices:
1) Keep the poor as they are: poor. Stop them from developing economically, so they can’t burn the CO2 which we, the rich folks, have been torching for decades now. Don’t let them have decent houses, clean water or high protein diets, because these things cost carbon, and we haven’t got it to spare. This solution would likely work but it seems to me to be highly immoral. I would hate to live in squalor, be hungry, or to not have healthcare. So I don’t want to espouse policies that depend on others having to live in a way I would not.
2) Move to a Global Single-child policy: The one-woman, one-child policy is the best way there is of controlling these effects. Policies which shift the birth rate equation down at all income levels are the ones most likely to achieve our environmental aims without having to inflict misery and suffering on our own species, or on others. It would take a policy step-shift in thinking to address these problems, but as far as I am concerned, when I look at the numbers, I am certain that this is the only reasonable policy solution there is. We can start by asking religious leaders like Pope Francis to change their messaging around birth control.
And, of course, we can stop thinking of demographic change in the West as a ‘problem’. It isn’t a problem, it’s the start of the only real solution.
NOTE: I am including the full set of data which I used to do all calculations. I welcome any corrections or suggestions for improving the model.
As with birth rates, we use data for 4 categories of countries from 1990 to 2015 (100 observations total). We have two explanatory variables, AGE and Y, where AGE is defined as the percentage of the population aged over 65 and Y is per capita GDP.
After eyeballing the scattergrams, we test the following functional form:
d = (minY^a)/Y^a * (1/AGE^g)
Where minY is the constant equal to the smallest value of Y in the series.
Logarithmic transformation gives:
ln(d) = ln(minY^a) – a*ln(Y) – g*ln(AGE)
which we test on the data using OLS. Here are the results:
Adjusted R square: 75.191
Intercept coefficient: 7.37384
Y coefficient: -1.01444
AGE coefficient: 2.0097
The estimated intercept is a good, but not perfect, approximation of ln(minY^a)
Here are the fitted against actual values of the scattergram for death rate against per capita GDP:
While the results are not as good as with the birth rates calculations, it is nevertheless a good enough fit and the explanatory variables have a strong enough confidence factor to be usable in our estimations.