Friday, June 11, 2010

Dissecting those ‘overpopulation’ numbers: Part Two: the perils of per capita

Continuing our examination of the misuse (deliberate or not) of numbers and statistics by advocates of the “too many people” explanation of environmental destruction.

Part One showed that the global population numbers frequently quoted by populationist authors conceal immense differences between countries. The principal conclusion was that “CO2 emissions are a problem of rich countries, not poor ones.” This means that programs that seek to mitigate global warming by reducing birth rates in poor countries are at best misguided.

Now we look at the misuse of “per capita” statistics. [1]

See also Part One: Population Where?

In an official briefing, the largest and most influential populationist group in the world explains why it favors a “population-based climate strategy”:

“The most effective national and global climate change strategy is limiting the size of the population. … A non-existent person has no environmental footprint: the emissions вЂsaving’ is instant and total.

“Given an 80-year lifespan and annual per capita emissions (2006) of 9.3 tonnes of CO2 … each Briton вЂforegone’ – each addition to the population that does not take place – saves 744 tonnes of CO2.”

The UK-based Optimum Population Trust (OPT) goes on to quantify the lifetime saving from preventing one birth at £30,000 – a “nine million per cent” return on a 35-pence investment in condoms.[2]

We might view that as a feeble attempt at humor, had OPT not published a clearly serious study in August 2009, purporting to prove that birth control is the most cost-effective way to reduce carbon emissions. The methodology involved forecasting the number of unwanted births that might be eliminated between now and 2050 if modern birth control were universally available – and then multiplying the number of non-people by the current per capita emission rates in the countries they wouldn’t be born in. Result – 34 fewer gigatonnes of CO2, at a cost of only $7/tonne.[3]

Prominent US populationist Lester Brown made a similar calculation to confidently predict that if the world’s population by 2050 matches the UN’s “low” projection instead of the “medium” projection, we will reduce our energy needs by the equivalent of 2,792 million tons of oil. He arrived at that improbably precise figure by multiplying the difference between the two population projections by per capita energy use.[4]

The source

OPT, Brown, and the many others who make similar use of per capita emissions figures, are using an argument that seems to have originated with Paul and Anne Ehrlich, for forty years the world’s most persistent and influential promoters of populationism. Their 1968 book The Population Bomb is the best-selling book on environmental issues of all time, and to this day it’s rare to find a magazine article on population that doesn’t cite or interview them.

In 1970, the Ehrlichs published a college-level text book whose central premise was set out in the Introduction: “The explosive growth of the human population is the most significant terrestrial event of the past million millennia. … Spaceship Earth is now filled to capacity or beyond.” [5]

They tried to express their views mathematically:

“It is axiomatic that the total impact of a society on the ecosystem can be expressed by the relation

I = P*F

“Where I is the total impact, P is the population size, and F is the impact per capita.” [6]

For F, you can substitute “per capita emissions” or “ecological footprint” or any other term that purports to measure individuals’ effect on the environment. Then, according to the formula, one person’s impact will be 1 times F, two people’s impact will be 2 times F, and so on.

Although longer versions of the formula have been proposed, this one is still widely used, both in narratives like the OPT statement quoted above, and in formal demographic studies. [7]

The median isn’t the message

At first glance, I = P*F appears obvious and undeniable – that’s why it remains central to populationist arguments. But its apparent simplicity conceals a common error: it mistakes an abstraction (an average or median) for concrete reality.

To illustrate this, we can turn to another field entirely. In 1982, the noted evolutionary biologist Stephen Jay Gould discussed his reaction to learning that he had a type of cancer for which the median survival time was just 8 months. His article was titled, profoundly, “The Median Isn’t the Message.”

“We still carry the historical baggage of a Platonic heritage … with its emphasis in clear distinctions and separated immutable entities, leads us to view statistical measures of central tendency wrongly, indeed opposite to the appropriate interpretation in our actual world of variation, shadings, and continua.

“In short, we view means and medians as the hard вЂrealities,’ and the variation that permits their calculation as a set of transient and imperfect measurements of this hidden essence. If the median is the reality and variation around the median just a device for its calculation, the вЂI will probably be dead in eight months’ may pass as a reasonable interpretation.

“But all evolutionary biologists know that variation itself is nature’s only irreducible essence. Variation is the hard reality, not a set of imperfect measures for a central tendency. Means and medians are the abstractions.”

Instead of despairing, Gould set out to learn how survival times varied concretely, and to find how he could maximize his chances of living longer than the “median.” In fact, he lived 20 more years. [8]

The populationist equivalent of “The median survival time for people with my kind of cancer is 8 months, so in 8 months I’ll be dead,” is: “Canada’s per capita emission rate is 17.2 tonnes, so reducing the population by one will cut emissions by 17.2 tonnes.”

Measures such as “emissions per capita” or “ecological footprint” are created by dividing a country’s total population into the total emissions generated in that country, or the total resources used there, or something similar. For comparing countries, a per capita figure can be more meaningful than a total – for example, the usual media assertion that China is the largest emitter of greenhouse gases conceals the fact that China has over four times the population of the number two country, the USA.

But per capita statistics can be very misleading if we forget that they are abstractions and treat them as concrete measures of the impact of individuals on the environment.

Just as Gould’s median survival time concealed a great range of possible outcomes, so per capita emission rates (or any other form of pollution per capita) conceal great variations in the pollution that can actually be attributed to specific persons or groups.

What’s included?

Perhaps the most obvious problem is that a large proportion of green house gas emissions are not caused by individual behavior at all. Simon Butler and I made this point in a recent article about claims that reducing immigration will reduce emissions.

“most emissions are caused by industrial and other processes over which individuals have no control.

“In Canada, for example, no change in the number of immigrants will have any effect on the oil extraction industry at the Alberta Tar Sands, described by George Monbiot as вЂthe world’s biggest single industrial source of carbon emissions.’

“Reducing immigration to the United States will have no effect whatsoever on the massive military spending – up 50% in the past decade – which ensures that the Pentagon is the world’s biggest consumer of oil. …

“Closing Australia’s borders would have had no effect on the climate denial policies of the previous Liberal Party government, or on the current Labor government’s determination to continue Australia’s role as вЂthe world’s largest вЂcoal mule.’”[9]

Reducing immigration won’t change any of those things, nor will reducing the population by other means. In fact, reducing population is more likely to have the paradoxical effect of increasing per capita emissions – proving once again that this statistic is an abstraction.

So at the very least, before claiming to predict the emission saving that will result from preventing one or a hundred or a million births, OPT should calculate a per capita figure that only includes emissions that are caused directly by individuals.

Wealth Matters

But even that wouldn’t make per capita emissions a useful guide to the effect of increasing or decreasing population. It may improve the numerator (emissions), but it still uses a denominator (population) that, as Marx wrote, is itself an abstraction.

According to an old joke, if you stand with one foot in the oven and the other in the freezer, on average you’re reasonably comfortable. That’s exactly the problem with “per capita” – it’s an average that conceals inequalities within the population.

A study published by the Canadian Centre for Policy Alternatives illustrates this. Drawing on path-breaking research by a former Director at Statistics Canada, Size Matters “finds that the ecological footprint of high-income Canadian households is substantially greater than that of everyone else.”

“The richest 10% of Canadian households are leaving behind an ecological footprint of 12.4 hectares per capita. To put that finding in context, their per capita ecological footprint is 66% higher than the national average. …

“The ecological footprint of the richest 10% of Canadians is nearly two-and-a-half times that of the poorest 10%.” [10]

The average ecological footprint increases gradually from 5.03 hectares in the poorest group to 8.87 hectares for those just below the top – and then it jumps by 40% to 12.42 hectares for those in the top 10%. As a CCPA news release announcing the report says, “When it comes to environmental impact, it really is a case of the rich and the rest of us.” [11]

Canada is not unique. A survey of household GHG emissions in India found even more extreme disparities – “the highest income group in India, constituting merely 1% of the population, emits four and a half times as much CO2 as the lowest income group consisting of 38% of the population.” A presentation accompanying that study concluded:

“There is no average India but a major divide between the emissions between the CO2 emissions of the rich and the poor.” [12]

Global Inequality vs Per Capita Pollution

In The Political Economy of the Environment, James K. Boyce considers the impact of global inequality on the environment, comparing the wealth not of countries but of people. He cites 1988 UN figures showing that the richest 20% of individuals in the world had average incomes of $22,800 each, while the poorest 20% averaged $163 each – a 140 to 1 ratio. He concludes:

“The amount of environmental degradation associated with a dollar’s worth of production and consumption is likely to vary across individuals and countries.

“Whether degradation per dollar is higher for poor people or for rich people is a question on which little systematic evidence has been collected. If the amount of degradation per dollar were roughly the same for both groups, the richest 20 percent of the world’s people would account for 140 times as much environmental degradation as the poorest 20 percent.

“Put differently, the total amount of degradation for which the poorest fifth is responsible could equal that for which the richest fifth is responsible only if the degradation per dollar for the poor were 140 times greater – a rather implausible suggestion.

“This simple comparison suggests that environmental degradation driven by the economic activities of the rich is likely to surpass, by a substantial margin, that driven by the economic activities of the poor.” [13]

The gap between rich and poor has expanded since 1988. In 2005 Thomas Pogge of Columbia University reported that the income ratio between the richest 10% of the world’s population and the poorest 10% was 320 to 1.[14] This makes Boyce’s conclusion even more credible today.

This gap is, inevitably, reflected in access to birth control: According to leaders of International Planned Parenthood, “In every country for which we have data, women in the highest quintile [richest 20%] have much better access to contraceptives when compared with the lowest quintile.” [15]

Despite this, groups such as Optimum Population Trust and the Sierra Club’s Population Justice Project claim that future Third World emissions can be significantly reduced by meeting “unmet needs” for birth control – that is, by providing reproductive health services to the world’s poorest people, who have the least access to birth control, who cause the least environmental damage today, and whose children are the least likely to contribute to increased emissions in coming decades.

As noted demographer Wolfgang Lutz has pointed out:

“within each country the rich have fewer children and emit significantly more than the poor. India, for example, has a per capita carbon emission of only 0.21 tons. Although this is one of the lowest in the world there is every reason to assume that the richest 10 percent in India emit at least 10 times more than the bulk of the population and that the expected future population growth of India comes almost entirely from the poor segments of the population. If this is true, the actual impact of population growth on carbon emission will be much less than national averages would imply.” [16]

We can restate his final sentence:

Reducing the number of babies born to poor women in India will have much less actual impact on carbon emission than national averages (per capita emission numbers) would imply.


A 2009 study of the impact of global inequality on green house gas (GHG) emission levels, by Dr. David Satterthwaite of the UK’s International Institute for Environment and Development, found that “a significant proportion of the world’s urban (and rural) populations have consumption levels that are so low that they contribute little or nothing to such emissions.”

Among his conclusions:

“It is very misleading to discuss responsibility for GHG emissions per person using national averages because of the very large differences in per capita emissions within each nation between the highest-income and lowest income groups – perhaps a 100-fold or more difference between GHG emissions per person if we could compare the wealthiest 1 per cent and the poorest 1 per cent in many nations …

“If GHG emissions were allocated to people (not nations) on the basis of the contribution of their consumption to GHG emissions, it is likely that the wealthiest one-fifth of the world’s population would account for more than 80 per cent of all GHG emissions (they have more than 80 per cent of the world’s income) and an even higher proportion of historical contributions to GHG emissions. The consumption of the one fifth of the world’s population with the lowest income levels may account for only around 1 per cent of all GHG emissions.” [17]

Part One of this article dissected the global population numbers that populationists so often cite, and concluded that CO2 emissions are a problem of rich countries, most of which have low birth rates, not poor ones with high birth rates. This means that the often-claimed correlation between global population growth and global warming is an illusion.

In Part Two, we’ve seen that per capita numbers, which are also used to support populationist conclusions, conceal substantial inequalities within populations and between people. When we dissect the numbers, we find that mostemissions and pollution cannot be attributed to the actions of individuals, and that insofar as some can be, they are mainly caused by rich people, not poor ones. This means that using ’per capita’ numbers or national averages to calculate the environmental impact of population changes will produce misleading results. (In short: I=P*F is wrong.)

So far, we have focused on quantitative issues – the differences in income or emissions between rich and poor. A unconvinced populationist could now reasonably respond: Yes it’s true that rich countries and rich people produce more emissions and other forms of pollution. But all people produce some, so any reduction in population will reduce the problem.

The ecosocialist response to that objection is that those overpopulation numbers don’t just conceal quantitative inequalities – they conceal qualitative differences of class and power.

We’ll discuss that subject in Part Three.

To be continued


Reference notes

[1] Judging by some of the responses I received to the first article in this series, I should point out that disagreeing with the analysis and solutions proposed by populationists is not the same as believing that the world’s resources are infinite, or that economic growth can continue forever, or that population size is irrelevant. As someone said, “Saying that I’m not at the South Pole does not mean that I am at the North Pole.”

[2] “A Population-Based Climate Strategy – An Optimum Population Trust Briefing.” Accessed May 26, 2010 at

[3] Thomas Wire. Fewer Emitters, Lower Emissions, Less Cost: Reducing Future Carbon Emissions By Investing in Family Planning. Accessed May 26, 2010 at

[4] Lester Brown et al. Beyond Malthus: Nineteen Dimensions of the Population Challenge. WW Norton, New York 1999. p. 47

[5] Paul R. Ehrlich and Anne H. Ehrlich. Population Resources Environment: Issues in Human Ecology. WH Freeman & Company, San Francisco. Second Edition. 1972. pp 1, 3

[6] Ibid, p. 252. This is an early version of the IPAT formula: Impact = Population times Affluence times Technology. Although it contains more terms, IPAT is effectively identical to – and has the same weaknesses as — its predecessor.

[7] See, for example, Brian O’Neill et al. Population and Climate Change. Cambridge University Press, 2001. p. 117.

[8] Stephen Jay Gould. “The Median Isn’t the Message.” Bully for Brontosaurus: Reflections in Natural History. WW Norton, New York, 1992. pp 473-478

[9] Ian Angus and Simon Butler. “Should Climate Activists Support Limits on Immigration?” Climate and Capitalism, January 24, 2010. Accessed May 27, 2010 at

[10] Hugh Mackenzie, Hans Messinger, and Rick Smith. Size Matters: Canada’s Ecological Footprint, By Income. Canadian Centre for Policy Alternatives, June 2008. p. 5. Accessed May 27, 2010 at

[11] Ibid.

[12] Hiding Behind the Poor. Greenpeace India Society, October 2007. Report, news releases and presentation accessed May 28, 2010 at

[13] James K. Boyce. The Political Economy of the Environment. Edward Elgar, Northampton, MA, 2002. p.6

[14] Summary of a discussion of the book Worlds Apart, sponsored by The Carnegie Endowment for International Peace, September 28, 2005. Accessed May 28, 2010 at

[15] Carmen Barroso and Steven W. Sinding, “Cairo: The Unfinished Revolution.” Laurie Mazur, ed., A Pivotal Moment: Population, Justice, and the Environmental Challenge. Island Press, Washington, D.C. 2010. p. 250

[16] Wolfgang Lutz. “World Population Trends: Global and Regional Interactions Between Population and Environment.” In Lourdes Arizpe et al, eds. Population and Environment: Rethinking the Debate. Boulder: Westview Press. 1994. p. 59

[17] David Satterthwaite, “The implications of population growth and urbanization for climate change.” Environment & Urbanization. 2007 Vol. 21(2) p. 545

Study: UT system brings in $2.5BResponse to “Who’s to blame for the population crisis?” in Mother Jones

No comments:

Post a Comment