Technology and humanity's impact on the environment

Note: I haven't forgotten about my promised followup to my article on Julian Simon, it will appear after this one.

Two people have commented on my article on the recent paper from the Optimum Population Trust, namely Martin Desvaux, the paper's author, and Tim Worstall, the blogger with an interest in economics, who's also been critical of Desvaux's paper.

The question at issue is whether technology increases or decreases humanity's impact on the environment and whether T in the I=PAT equation should be considered to be a multiplicative term or whether we should divide bed T. Note that I measures the human impact on the environment, P is the population and A is a measure of consumption (or affluence).

I have thus far sided with Tim Worstall's view that technology helps us reduce our impact on the environment, however whilst Devaux acknowledges that technology can do so, he argues that overall it has been a driver in increasing humanity's impact on the environment. From his comment:

I think it is safe to state that, since the industrial revolution, technology has enabled us to reduce infant mortality, increase food production and increase life expectancy, all of which have caused the largest population explosion in the history of humankind. I think we can also safely assume that in that arena T is greater than 1 as a result. Such progress was possible only because we could extract oil, coal and gas out of the ground in ever increasing quantities, transport it via road, rail, pipe and sea* all around the world which, I feel sure you will agree, has also had a T-greater-than-one impact on the environment. In addition, we get most of our fertilizers from hydrocarbon technology. Without fossil fuels and thereby electrical energy, medical advances would have been impossible, We would not have been able to develop (to mention those which immediately spring to mind): warm homes, fridges, leisure centres Olympic stadia, moon shots, as well as several billions of cars, millions of lorries, ships buses, railways, aircraft, agricultural machinery factories, processing plant …. with all the infrastructure of roads, ports, depots, etc that these entities require.

What Devaux is saying here is that the advances in technology associated with the industrial revolution and the fossil fuel economy have led to a huge increase in both population and affluence, which has led to an increase in humanity's impact on the environment, and thus we should consider T to be a multiplicative variable, with value greater than 1 in the I=PAT equation.

I think this is a misunderstanding of the role of T in the equation. Suppose we rewrite the equation to give us T in terms of the other three variables. We then get T=I/(PA). T is thus measuring the environmental impact per unit of consumption, where A is consumption per head of population and P is the size of the population. Stated in these terms, we can see that T does not measure "technology" per se (as I argued here) but rather measures a variable which technology can influence.

As Tim Worstall points out, Desvaux is double counting. Desvaux suggests that technological changes drove an increase in population and an increase in affluence, thus implying T should have a value > 1. The problem is that in the I=PAT equation, P and A already reflect those increases, thus to incorporate those increases in the value of T involves incorporating them twice!

This illustrates a weakness of the I=PAT equation. It treats P, A and T as independent variables when in fact there are feedbacks between them. But Desvaux is surely correct that technological advance enabled the large human population we now have and the levels of affluence we now see, and as I pointed out earlier, increases in affluence have led to a fall in birth rates in developed countries (and increasingly elsewhere) and thus to a slowing of population growth in recent decades.

But equally one can point out that without our technology, it simply wouldn't be possible to support over 6 billion people, and we would devastate the environment if we were to try doing so. So where does this leave us on whether technology increases or reduces our impact on the environment?

My view is this. Technology can do both. We employ technology because it makes things easier to do. It can do this in various ways:

• It can substitute for human labour. For example, a man who spends 30 mins walking to work each day might buy a car to reduce the journey time to 10 mins. Getting to work is now easier and more comfortable, but his environmental impact has increased and he will be using more resources. E.g. instead of moving one human to and from work, he's now moving one human plus a ton or two of metal and thus using more energy.
• It can enable us to do things we couldn't or wouldn't do before. E.g. the man who buys a car to shorten his journey to work now visits his relatives 60 miles away several times a year, where previously he would have used public transport and only gone once or twice at most. Again this increases the impact on the environment, but this time it is because the man is engaging in more activities than he used to. Another example is the summer holiday abroad that many people take which they could not do were it not for the advent of air travel.
  
• It can enable us to do things with fewer resources. Suppose the man who bought the car above, later on buys a new car with twice the fuel efficiency. His journeys to/from work and his relatives now have a lower environmental impact since less fuel is needed to power these journeys.
• It can enable us to tap new or previously unconsidered resources. For example, the internal combustion engine indicated that some black gooey liquid, commonly called oil, actually has its uses and advances in pumping and drilling technology enabled us to extract oil from previously inaccessible places. This of course had a considerable environmental impact, though supporting 6 billion people without such a concentrated portable energy source would likely have devastated the environment were it to be tried.
  

Whether overall technology will increase or reduce environmental damage depends on the choices we make. If the environmental damage becomes serious enough we will choose to mitigate it. If the cost of such damage can be internalised so that e.g. the polluter pays for his pollution, then technology will tend to develop in more environmentaly friendly ways. We should thus look at ways in which technology can reduce the value of T in the I=PAT equation.

The Commoner-Erhlich Equation

Further to my previous article, on the report from the Optimum Population Trust, I've been doing a bit of digging around on the I=PAT equation. Remember here that I is the measure of the impact of humanity on the environment and P is the population and A is a measure of affluence (or consumption). The question is what is T measuring? The OPT reports talks about T somehow measuring "technology".

Anyway according to Wikipedia, T is in fact humanity's ecological impact per unit of consumption. A is measured as consumption per capita. So by multiplying the population P by the consumption per capita A, you get total consumption, after which you multiply by T the total impact per unit consumption to get I, the total environmental impact of the population and its level of consumption.

Given this, it is clear Tim Worstall's criticism of the I=PAT equation, saying that we should divide by T, not multiply by it, is mis-placed. Mr Worstall is treating T as if it measures technological sophistication. I agree with him that technological advancement reduces our environmental impact, at least for a given standard of living and population size, but that is not what T is measuring here. Technological advancement allows us to e.g. use less energy and resources and/or reduce pollution per unit of consumption. Thus such advancement reduces the value of T. The question then is whether the equation is an adequate description of what's going on. It assumes independence of its variables and it also assumes the variables can be measured reasonably accurately. It seems to me both assumptions are questionable.

For example, there may be feedback loops between the variables that aren't catered for and it's not entirely clear how one would measure either "consumption" or "environmental impact" in a clear, accurate manner.

Technology, Affluence and the Optimum Population Trust

The Optimum Population Trust recently published a study which claims that Britain's optimal population is about 17 million people:

If the UK had to provide for itself from its own resources, it could support a population of only 17 million – 43 million less than its latest official population figure* - according to new research by the Optimum Population Trust.

Even if the UK dramatically improved its sustainability with a 60 per cent cut in carbon emissions by 2050 - the target set by the present Government - UK “overpopulation” would grow from 43 to 50 million, the research shows. This is because projected population growth of 17 million**, taking the country’s population to 77 million by 2050, would cancel out the sustainability benefits of carbon savings.

The sustainability of human populations: How many people can live on Earth? ***, published today (Monday February 18), is based on a new analysis of biological capacity and ecological footprinting data. It suggests that in 2003, the last year for which comprehensive data are available, total world population was 6.3 billion but the sustainable figure was 5.1 billion. Global overpopulation was thus 1.2 billion. (italics in original)

A 9-page report based on this study can be downloaded here. From pages 2 to 3:

Not surprisingly, the impact of this population growth on the environment since 1750 has been extensive. Now, not a day goes by without news of droughts, floods, famines, conflicts over resources, extinctions, and, in the last 20 years, the increasingly evident effects of global warming. This impact has been expressed in what has become known as the Commoner-Ehrlich Equation:

I = P x A x T.

This states that the impact (I) on the environment is directly proportional to the population size (P), the ‘affluence’ (A) (defined as the resources a population consumes and wastes) and technology (T) through which we (1) prolong life, (2) produce things more quickly and cheaply (thus feeding back into consumerism and affluence) and (3) grow food faster which feeds back into ‘population’. This equation thus neatly summarises the impact of humankind on the planet.

Note that it is assumed that technology is multiplicative factor that increases the human impact on the environment. Yet technology mitigates the impact we have on the environment by enabling more efficient use of resources and/or less polluting methods to be used.

It is technology that has enabled us to sustain the large population we currently have on earth, living longer and healthier than at any time in history. Remove the technology and the environment would be devastated as people desparately try to grow food and obtain water using methods that simply cannot sustain us. Indeed, based on similar points to mine above, Tim Worstall argues that we should divide by T rather than multiply. However reading further, it seems that T isn't measuring technological advancement, but rather the impact of technology on the environment:

Politicians, unsure what to do, offer solutions which include suggestions such as: develop fuel-efficient cars; change to efficient light bulbs; fly less; build renewable energy and nuclear power plant; increase mass transit systems; and plant trees. These solutions only address the reduction of the affluence and technology variables of the equation, but never the population variable.

Reducing impact by decreasing affluence (consumption) only partly addresses the problem since populations are growing faster than affluence – for example, in Africa. Technology, meanwhile, tends not to “decrease” at all. Whilst it can be used to reduce the impact of affluence, it is likely that its benefits in energy saving devices will be cancelled out by its disadvantages, as businesses continue to use it to maximise their economic growth via consumerism. So, realistically, impact will continue to rise since economic growth demands it. This is bad news since, as we will now see, human impact on the planet is already unsustainable. (italics in original)

Here the paper acknowledges that technology can in fact reduce the impact of humanity on the environment (though it argues that the drive to economic growth will then cancel this out). To retain T as a multiplicative variable, whilst acknowledging that it can reduce humanity's impact on the environment, one must consider it to be a measure of the impact of our technologies on the environment, rather than a measure of advancement. Technological advancement will thus tend to reduce T, and I'd suggest it has been doing so for centuries whilst increasing population and affluence have offset the reductions in impact it enabled.

An interesting point is that there is no mention in this study of one of the main findings in demography which is that increasing affluence has lead to a fall in birth rates resulting in slow population growth rates or even declining populations in rich countries. This implies that rising affluence may in fact help with the goal of slowing population growth, a finding that is at odds with the arguments presented on the OPT's paper.

I intend to return to other aspects of this paper in later posts.