Demographic Transition Through Increasing Energy:
Can It Work?

In my recent article on  the correlation of energy consumption and population I made the following claim: "..improving humanity's energy supply picture will not halt population growth.  It would certainly not curb it until the bulk of the world's population had achieved the per capita consumption of the developed world, where Total Fertility Rates have finally declined below replacement rates.  This is, of course, just a restatement of the Demographic Transition Model into energy terms, away from the commonly assumed proximate cause of industrialization."

I further claimed, based on some preliminary and overly general calculations, that it would take on the order of three times our current total primary energy output to stabilize the world population at around 10 billion people.   This article is an attempt at a more rigorous analysis of the relationship between energy and fertility rates, with a goal of  increasing the confidence of such a projection.

Demographic Transition

Demographic Transition is defined in Wikipedia as follows. "The term demographic transition is a theory describing a possible transition from high birth rates and death rates to low birth and death rates as part of the economic development of a country from a pre-industrial to an industrialized economy."  While demographic transition has never been experimentally verified, the effect has been observed often enough for the theory to have become well accepted (perhaps even a sacred cow?) in demographic studies.

The first question I needed to to resolve was whether the correlation between per capita energy consumption, Total Fertility Rate (TFR) and Life Expectancy (LE) is sufficiently strong to support the use of energy as a proxy for industrial development in the Demographic Transition Model (DTM).  To do this I first established the population, primary energy consumption (i.e. the aggregated consumption of all forms of energy), TFR and LE for as many countries as possible (1). The per capita energy consumption (expressed in "Tonnes of Oil Equivalent", or TOE per capita per year) was then calculated for each country.  Finally a pair of scatter graphs were constructed, one each for Energy vs. TFR and Energy vs. LE.  In both cases the countries were weighted by population.  The points on each graph were then matched with an Excel trend line chosen to give the best observed fit.

70 nations or regional aggregates are represented in the data set.  This number was established by the national breakouts given by BP in their Statistical Review workbook.

The following graphs are the result:

Enerby vs. TFR

Energy vs. Life Expectancy

In each case you can see a strong correlation, though the correlation with TFR is higher than with LE.  The trend lines are logarithmic, with the result that small changes in energy consumption have more of an effect in low energy consuming poor countries than in high energy consuming
rich ones, as you would expect.  Overall, the results pass my sniff test.  They seem to confirm that per capita energy consumption is an acceptable proxy for the vague and hard-to-measure variable of "industrialization" as it is used in the DTM.

The Application to Population Stabilization

From the demonstrated correlations we are now sure that a country's fertility drops as its energy consumption rises.  It should therefore be possible to quantify the amount of energy needed to stabilize the world's population.  That would mean supplying the world with sufficient energy to enable enough industrialization to bring the global TFR down from its current value of 2.8 to a level of 2.1 or less.

In order to determine what amount and distribution of additional energy might be needed to accomplish this, I first characterized the world's current population in terms of TFR and energy consumption.

I divided the world population at the "replacement" TFR of 2.1.  Interestingly, the split was quite uniform: about half the world population is over that dividing line, and about half is under it.

The per capita energy consumption of each group was calculated.   To establish an energy target, I also calculated the per-capita consumption of the group of countries with a TFR from 1.75 to 2.1, including both China and the USA.   This characterization is shown in the following tables.

The World
Population (Millions)
Energy Consumption (MTOE/yr)
Average Per Capita Energy Consumption (TOE/yr)

The Rich Half
Population (Millions) 3,066
Energy Consumption (MTOE/yr) 8,668
Average Per Capita Energy Consumption (TOE/yr 2.83

The Poor Half
Population (Millions)
Energy Consumption (MTOE/yr) 1,869
Average Per Capita Energy Consumption (TOE/yr) 0.55

The Countries Around TFR 2.0
Population (Millions) 2,070
Energy Consumption (MTOE/yr) 4,815
Average Per Capita Energy Consumption (TOE/yr) 2.3

The theory we are exploring predicts that if we could bring the energy consumption of the poor half of the world's population up to the level enjoyed by the countries with a TFR around 2.0, that there is a good chance their populations would stop growing.  We of course don't need to worry about the rich half of the world, since its population has already stopped growing.  The above numbers tell us is that we would need to increase the energy available to the poor half of the world to about four times its current level, to bring the average consumption from  0.55 TOE/yr to a level of 2.3 TOE/yr.

One problem with this approach is that the global energy market is more or less a free market.  This makes it very difficult to "direct" energy supplies to poor countries, since they have less funds to buy the energy compared to rich nations.  For now, we will ignore this inconvenient truth to some extent (and in the end it won't really matter).

Of course it's not possible for the world to suddenly produce a huge amount of new energy.  Energy production typically rises slowly over time, and during that time the population keeps expanding, thus raising the total energy needs ever higher.

In order to limit the effort required, I investigated only two scenarios.  The first is a hybrid that falls somewhere between a pure directed world economy and a pure free market.  I envisioned some (unidentified) central agency that could direct a large fraction of new energy production to poor nations, but would still allow rich nations to expand their energy consumption somewhat.  The second scenario is more of a free-market system, in which new energy is distributed evenly to all countries, and enough new energy is produced to lift the poor half up to the required 2.3 TOE/cap/yr.  I didn't consider the effects of a true free market system in which energy is supplied preferentially to those who can pay for it.  That was because it became clear from the first two scenarios that the requirement for  higher energy production entailed by such a distribution scheme was utterly unrealistic.

In each case I targeted to arrive at a stable population in 50 years, in 2057.  I also postulated a gradually declining net birth rate down to a stable replacement level at that time.  This gave a final population figure of 9 billion in each case. The difference was in the amount of extra energy required to get there.

Scenario 1

The increase in energy availability was set at four times current consumption for the poor nations (0.55 to 2.3 TOE/yr), and twice current consumption for the rich ones.  The global average energy consumption target for this scenario was then calculated as 3.8 TOE/cap/yr.  Given our current world level of 1.67, this would require an increase of  just over two times.  However, as I mentioned above, as we're ramping up our energy production, the population keeps on growing.

It turns out that the target of 3.8 can be achieved in 50 years with an average growth in the energy supply of 2.4% per year.  This is similar to the growth rates we have actually achieved over the last ten years, so at first it may not appear too difficult.  However, this scenario requires us to do this every year for fifty years.  That seems decidedly unrealistic given Peak Oil constraints, the opposition to nuclear power, the risks from global warming if coal is used, and the relative infancy of the wind, solar and tidal power industries.

The final result for this scenario is that in 2057 we have a world with a stable population of nine billion, using over three times the energy we are today. This is not an encouraging outcome, especially as it is the result of an unrealistic requirement for a directed global energy economy.

Scenario 2

In this scenario, new energy was assumed to be distributed uniformly to all countries.  Again it's unrealistic from a market perspective, but at least it requires somewhat less global direction than Scenario 1.  The entire world's energy consumption was allowed to rise fourfold, to make sure that the poor half made it to the stability point of TFR 2.0.  The global average energy consumption target for this scenario was calculated as 6.5 TOE/cap/yr. All the assumptions about population growth rates were retained from Scenario 1, though it's probable that the greater rise in energy would affect them to some extent.

In this case the target consumption was reached in 50 years through an average growth in the energy supply of 3.3% per year.  This is about 50% higher than the recent historical growth of energy supplies.  The bad news is that by the end of those 50 years, we would have to be producing over five times the energy we are today.  If the outcome of Scenario 1 was disappointing, this one seems frankly impossible.

In the face of this result, it seemed unnecessary to explore the energy requirements under a truly free global energy market.  Given the differential buying power of rich and poor nations, it seems likely that an order of magnitude increase in the world's energy supply would be needed to accomplish the goal.


This entire thought experiment revolves around the use of the DTM as the mechanism for stabilizing our population.  Given the results, it seems obvious that even if  that mechanism worked, humanity could not afford it.  The environmental impact of quintupling our energy production is simply unimaginable.  As well, it is unclear where the energy would come from, and the costs of building new production facilities plus the required infrastructure would probably break humanity's bank.

There is one last over-riding consideration that makes the endeavour unrealistic.  Even if we could do it, we would end up with 9 billion people all wanting to live a modern middle-class lifestyle, with all the activity and consumption that implies.  Regardless of whether they have a "right" to such a lifestyle or not, the ecosystem of our planet would simply not survive such an onslaught of resource depletion and waste generation.  We're barely hanging on as it is, and the expansion of the human footprint this would entail would probably ring the death knell for most of the biosphere.

As a result, we must look to other mechanisms to lower our growth.  The example of Kerala, involving high levels of education and empowerment of women along with universal access to family planning resources, shows that it is possible in some limited situations to achieve a good quality of life in a low-energy society.  This option should be vigorously pursued.

Is there a realistic way out of our population trap? Can we influence the present course of events enough to avoid an ecological and population crash?  Unfortunately, I have not yet found anything that convinces (or even suggests) that the answer is "Yes.".

August, 2007

  1. Data on national energy consumption came from the BP Statistical Review of World Energy 2007. Population figures came from the CIA World Factbook. TFR figures and Life Expectancy data came from Wikipedia.
© Copyright 2007, Paul Chefurka

This article may be reproduced in whole or in part for the purpose of research, education or other fair use, provided the nature and character of the work is maintained and credit is given to the author by the inclusion in the reproduction of his name and/or an electronic link to the article on the author's web site.  The right of commercial reproduction is reserved.