Why Julian Simon is both right and wrong

[This is the followup to my earlier article about Julian Simon.]

Now I'll restate this line of thought into a theory that will appear again and again in the book: More people, and increased income, cause resources to become more scarce in the short run. Heightened scarcity causes prices to rise. The higher prices present opportunity, and prompt inventors and entrepreneurs to search for solutions. Many fail in the search, at cost to themselves. But in a free society, solutions are eventually found. And in the long run the new developments leave us better off than if the problems had not arisen. That is, prices eventually become lower than before the increased scarcity occurred. (From The Ultimate Resource 2, Chapter 3)

This is perhaps the key passage in Chapter 3 of Simon's book. It highlights a key part of the drive that has led humanity to develop a dizzying array of technologies, achieve the longest lifespans, the most comfortable lifestyles and the healthiest populations in history. It also illustrates why Paul Erhlich lost his infamous bet with Julian Simon.

Unfortunately, the copy of the book on the internet I've been using seems to have Chapter 3 cut short, so I don't have the reasoning there that takes Simon from the problems of defining "natural resources" discussed in Chapter 2 and the passage above to his conclusion that resources are not finite. However, an article of his published at the Cato Institute, does shed some light:

...the term "finite" is not only inappropriate, it is downright misleading when applied to natural resources. The mathematical definition of "finite" is quite different from a useful economic definition.

For instance, the quantity of services we obtain from copper should not be considered "economically" finite because there is no way of counting them appropriately. We should also consider the possibilities of using copper more efficiently, of creating copper or its economic equivalent from other materials, of recycling copper or even obtaining copper from sources beyond planet Earth.

Therefore, a working definition of the total services that we could obtain from copper now or in the future is impossible to construct. (emphasis added)

There is also his reply to critics in which he says:

Finiteness by itself is not testable, except insofar as the fact that no one is able to state the absolute size of the relevant system (our cosmos) demonstrates the absence of finiteness in its dictionary sense. But the relevant evidence we have available - decreasing prices and increasing substitutability - is not what one would expect from a finite system. (emphasis added)

And:

Nothing I have written is intended to suggest that during any particular period there may not be too much use of any resource, renewable or non-renewable; indeed, I expect temporary overuses (for example, overuse of forest resources in various countries in various centuries) just as I expect boom-and-bust cycles in all other human endeavors. But this is a matter of management and adjustment in dealing with, and riding out, the ups and downs, rather than a matter of ultimate finiteness.(emphasis added)

From this I posit that Simon's argument can be boiled down to the following:

• As reserves of resources run down, the resulting price rises spur the search for new sources of them, for more efficient ways of using them and for ways of substituting other resources for them.
• The long run trend (for centuries) has been for the price of resources to continue falling. Temporary shortages have often led to discoveries that leave humanity better off than before those shortages occur.
• We do not know, ultimately, what resources are available to humanity in the long run. All we know is what resources are available now/in the forseeable future, given current technology.
• We don't know whether the universe is finite or not, and we cannot thus state that the resources available to us are finite. The long run trend of falling prices and greater abundance of resources seems at odds with the assumption of finiteness.
• Since we do not know what resources will be ultimately available to us, we cannot say they are finite in any meaningful sense.

There are several problems here:

1. We do know that the earth is finite. This is an incontrovertable fact. There is a finite amount of energy reaching earth from the sun each year, and a finite amount of matter falling to earth each year from outer space. Until we can exploit extra terrestrial resources at least as easily as we currently exploit the resources on earth, i.e. until we can escape the confines of earth as easily as we can escape the confines of a continent, this really does limit how many people the earth can support and the standard of living those people can enjoy. That seems unlikely to happen for at least a century --- on that timescale the most I'd expect is colonies on the moon and a manned trip to mars.
   
2. The trend for falling resource costs is a matter of a few centuries -- this is a short time compared to (a) recorded history (b) the existence of humanity. We know that civilisations in the past have thrived and then collapsed. It seems likely that some of them died because of resource shortages.
3. For the process of resource discovery and creation to keep us from "running out", it must produce new resources at or above the rate at which we consume them. If we're to rely on this process to prevent disaster, we must therefore posit that there will always be sufficient resources that can be reached via the process within the timescale required to stave off disaster, at every point in time. It seems to me unlikely that this can be guaranteed.
   

Simon is correct to highlight the existence of the process of resource discovery and creation, and at a highly abstract level he is even right that we don't know whether the resources ultimately available to humanity are finite or not. But the process is not automatic, and even when running efficiently, it is not guaranteed to provide us with all the resources we might need at a given point in time.

To act as if resources are infinite, when we know that running out is a real possibility and when even our most advanced science and technology tells us we can do no more than an exploratory flight to our nearest planetary neighbour (let alone colonise it, terraform it or get there in the sort of timescale we can travel to other continents) would be irresponsible.

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.

Julian Simon, the limits to growth and infinite resources

The Earth's physical resources are finite. We threaten our future if we try to live beyond those means, so we must build a sustainable society that guarantees our long-term future.

The above quotation comes from a document describing the philosophical basis of the Green Party. The late economist, Professor Julian Simon, rejected this view in his book, The Ultimate Resource 2. To quote from Chapter 3 (entitled "Can the supply of natural resources - especially energy - really be infinite? Yes!"):

Chapter 2 showed that natural resources, properly defined, cannot be measured. Here I draw the logical conclusion: Natural resources are not finite. Yes, you read correctly.

So here we have a deceased but influential economist claiming that resources are infinite! Note the "properly defined" bit above. The following paragraph from the summary of Chapter2, illustrates the thinking here:

Material-technical forecasts of resource exhaustion often go wrong for two reasons. (1) No matter how closely defined, the physical quantity of a resource in the earth is not known at any time, because resources are sought and found only as they are needed; an example is the increase in the known supplies of such resources as copper, as shown in table 2-1 and figure 2-1. (2) Even if the physical quantities of particular closely defined natural resources were known, such measurements would not be economically meaningful, because we have the capacity to develop additional ways to meet our needs - for example, by using fiber optics instead of copper wiring, by developing new ways to exploit low grades of copper ore previously thought not usable, and by developing new energy sources such as nuclear power to help produce copper, perhaps by extracting it from sea water. Thus the existing "inventory" of natural resources is operationally misleading; physical measurements do not define what we will be able to use as future supplies.

What Simon has demonstrated is that it is hard to measure what resources are available in the earth, that we don't know what future means of providing those resources or substituting for them will become available and it is thus hard to define what resources the earth can/will ultimately provide to humanity.

However he has not demonstrated that the earth can provide us with "infinite" resources, merely that we do not know what resources it could ultimately provide us with. However Chapter 3 is where he attempts to demonstrate that resources are in fact "infinite". I shall tackle the reasoning there in my next article.

One might ask why I am bothering with this. The answer is that, whilst Simon is wrong about getting "infinite" resources from the earth, he has sound arguments to make about the economics of resource usage/scarcity and I believe he goes wrong in an "interesting" way. Understanding where he goes wrong can help understand what the real situation is regarding "the limits to growth".

Blog topics

I'm currently adding labels to old posts produced before labels were added to blogger.com.

The result so far is an increased list of topics in the "Topics" side bar (this used to be labelled "Labels"), plus an increased number of articles indexed by the labels. My aim is to get rid of the (renamed) "Google Searches" side bar that I used a substitute for labels. I won't necessarily catch all articles, but hopefully it will make it easier for readers to look up the material that interests them in this blog.

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.

Yusuf Al-Qaradawi and freedom of speech

Inayat Bunglawala, writing in the Guardian, claims that the British government's refusal to give Yusuf Al-Qaradawi a visa, thus refusing permission to enter the country, violates Qaradawi's freedom of speech:

Gordon Brown's government has finally caved in to the noisy mob who have been angrily demanding that the elderly Islamic preacher, Yusuf al-Qaradawi, should be refused a visa to come to the UK for medical treatment.

Well, so much for free speech. You will recall that during the Satanic Verses and the Danish Cartoons row, British Muslims were repeatedly lectured to about the need to adapt to western notions of free speech. You may not like what is written or drawn, we were told, but as long as it does not break the law, you need to learn to put up with it.

The problem with this argument is that Mr Qaradawi's freedom of speech has not in fact been curtailed. His words have not been banned from the media or the internet, he can continue giving interviews, making speeches, etc. It's just he's been refused permission to enter the country, which is no more of a violation of his freedom of speech than if I were to refuse him entry to my house.

The point is that freedom of speech is the right to express your views with your own resources, or resources you otherwise have permission to use, to anyone willing to listen. Freedom of speech does not give me the right to enter your house without your permission. Similarly it does not give a non-citizen the right to enter a country, whether he wishes to do so in order to spread his views or simply to have a holiday. The non-citizen must get permission from the country's government to do so (said government exercising this power on behalf of the people of that country).

In practice permission is often granted by default, assuming you apply/arrive through legal channels. But governments have always had the power to refuse permission, a power which the are supposed to exercise in defence of the country concerned (e.g. to repel foreign invaders or anyone else who poses a risk to that country's population). In this case, the British government has decided Al-Qaradawi poses some sort of threat. Whether they are right in that decision is a separate matter from any alleged violation of freedom of speech.

Mr. Bunglawala is confusing freedom of speech with the right to be provided with a platform of one's choice in a location of one's choice. No one has that right.