Contents - Previous - Next

This is the old United Nations University website. Visit the new site at


It is again useful to start with a simple numerical example, still using the simplified version of the Ehrlich equation. Suppose that a country is composed of two groups, A and B. with the characteristics at date 1 and 2 as shown in Table 5.2.

Table 5.2 Population groups and environmental impact

Date 1 Group A Group B Total
Population 90 10 100
Total impact 0 10 10
Impact per person 0 1 0.1
Date 2 Group A Group B Total
Population 190 10 200
Total impact 0 20 20
Impact per person 0 2 0.1

If decomposition of the country between group A and group B is impossible, that is if the levels of population and consumption are only observable at the country level, than it will appear that the contribution of population growth to the increase in environmental impact is 100 per cent (whatever the definition used), since there is no increase of the average impact per person. But this result is clearly not satisfactory since the group which grows in size has no impact on the environment.

The problem is sometimes acknowledged, for instance by Harrison (1992: 313), and such studies disaggregate between developing and developed countries or between continents. But others are much less careful. Thus, in an UNFPA publication, Myers (1991) could argue that population growth is responsible for about two thirds of the increase in the emissions of carbon dioxide between 1950 and 1985, simply because during the period, the emissions of carbon dioxide grew by 3.1 per cent a year and world population by 1.9 per cent. But since the population of developing countries grows fast in size but consumes very little resources, whereas the reverse is true for developed countries, there is a problem of heterogeneity. A simple disaggregation between developed and developing countries reveals the oversight. In doing so, Harrison found that population growth would account for 41 per cent of the increase in emissions between 1960 and 1988 using the same definition as Myers (definition 3).

Myers is not an isolated example. Pearce (1991) makes a similar assumption while assessing the contribution of population growth to the increase of energy consumption between 1960 and 1984 using the first definition (ie consumption per person of reference is that of 1960). At the world level, when calculated directly as Pearce does, this contribution would be 46 per cent. Simply using the disaggregated results by continent and weighting them by the respective increases of consumption yields a very different result indeed: 23 per cent.

Disaggregating between North and South yields other quite paradoxical results. When we speak about population growth, it is 'obvious' that we speak about population growth in the South. This is indeed where 90 per cent of population growth took place during the last three decades. Yet, according to Harrison's calculation (using definition 3), population growth in the North would have accounted for a greater share of the global increase in the emission of CO2 than population growth in the South: 22 as against 17 per cent. This is because the consumption per person factor is so much greater in the North, a point indeed stressed by Ehrlich and Ehrlich (1992: 58). Thus, with the same definition and the same data, if we speak about the contribution of population growth at the world level, we find 64 per cent. But if we speak about the contribution of population growth taking place in developing countries, we find only 17 per cent!

This problem of aggregation is particularly critical because consumption per person and fertility rates are negatively correlated around the world. That is, because the poor tend to have more children and the rich less, in any aggregate, the equation will overestimate the contribution of population growth. This is true at the world level, as revealed in the proceeding examples. But it is also true at a continent or country level, although none of the studies reviewed take the issue into account. The 17 per cent above is therefore already a large overestimation of the environmental impact of population growth among the poor in developing countries. If we could statistically distinguish between the 10 per cent richest in developing countries, and the rest of the population, then we would certainly find that 80 per cent of world population growth has contributed less than 10 per cent of the increase in total emissions of CO2. Table 5.3 illustrates how misleading are results not taking into account the issue of heterogeneity. Finally, it must be noted that heterogeneity in the technologies used introduces yet another bias in the calculation of the different contributions.

We must however be cautious while disaggregating the data, when statistically feasible, to take into account use of natural resources for exports, as not doing so might give very misleading results. For instance, looking at changes in farms' areas (area of arable and permanent crops) between 1961 and 1985, Harrison found that population growth would account for 72 per cent of the change in developing countries. The technological factor, here the inverse of yield, would have had a negative impact, that is it would have tended towards reducing the area under cultivation by 2.6 per cent per year. Put another way, the demand for land has risen by 3.2 per cent per year (population growth 2.3 per cent, per capita consumption growth 0.9 per cent), and this demand was met by a greater productivity (2.6 per cent) and by an extension of land area under cultivation (0.6 per cent).

Table 5.3 World carbon dioxode emissions, 1960 1988: contribution of population growth

  Contribution of
population growth (%)
% of total
population growth
Crude world calculation 64 100
Disaggregated world calculation 41 100
LDC contribution to DWG 17 89
Contribution of the 'poor' in LDC <10? 80  

Note: Definition 3 used for calculation of respective shares.

These results are of course open to the critique of heterogeneity, the more so as access to land is particularly inegalitarian. Inequality in land distribution is therefore not considered as a potential factor of environmental degradation. But even more importantly, the whole effect of commercialization and the use of land for exports to developed countries is not taken into account, although this is held as a main cause of environmental degradation (Lappé and Collins 1980; Repetto and Holmes 1983). For instance, commercial logging for the export market has been a main force behind deforestation; another important force, notably in Brazil, has been the establishment of cattle ranches, again turned towards exports. But this demand is not taken into account because of the separation between developing and developed countries.

Are the Variables Independent?

Applying the Ehrlich equation supposes that the three variables P. T and C' are independent. That is, it is assumed that an increase in P (population growth) will have no effect on the technology used or on the level of consumption per person. This is a very strong assumption, and one that the authors have adopted perhaps without stopping to gauge its significance. Yet it openly conflicts with a number of theories on the consequences of population growth.

In their seminal article Ehrlich and Holdren (1971) had discussed why, in the case of industrialized countries, an increase in population could have a positive effect on T. thereby reinforcing its impact on the environment. In general, and notably in developing countries, the inverse might well be true. For instance, it has been argued that population growth was itself a main cause of technological innovation (Boserup 1965). A higher density of population not only permits the implemention of new methods of production, such as a greater division of labour, but also creates the need for new methods when the environmental constraint is binding. For Boserup, it is therefore both a condition and an incentive for innovation. A number of studies on the subject have shown that greater population pressure led to the adoption of labour-intensive farming methods, without necessarily degrading the environment (Hayami and Ruttan 1971; Binswanger and Ruttan 1978). In other cases, the environmental constraint may be stringent, and population growth will lead to a reduction of levels of consumption per person. In any case, each individual addition to the population brings with him or herself a new capacity to innovate, and this will possibly more than compensate what he or she will take from the world, as forcefully argued by Simon (1981).

It has also been argued that population growth could have a negative effect on consumption per capita. This is in fact a central argument put forward by those who view population growth as a problem. 'A population problem exists', argues Demeny (1986: 481), 'when my preference for children diminishes your access to steak.. We have a population problem, in other words, when externalities are attached to demographic behaviours.' That is, although it may be in a couple's immediate interest to have many children, if all couples behave similarly, and because some resources are finite, eventually everybody will be worse off, a process often referred to as the 'poverty trap'. In fact, if all resources have been appropriated (whether held privately or in common), the impact of population growth on these resources will not be direct, but transmitted through the institutional framework. Taken to extremes, if the management of a resource is independent of the demand for this resource (I = constant), an increase in the size of the population will automatically lead to a decrease in consumption per capita if T is assumed fixed. Not taking into account this sort of mechanism will lead again to an overestimation of the impact of population growth on the environment.

Conversely, and according to demographic transition theory, an increase in consumption per capita will possibly have an impact on fertility rates, and thus on the level of population in the long term. Under this argument, the contribution of the consumption factor would be overestimated.

The point is that population growth can be accommodated in different ways. How it is accommodated, and therefore what will be the consequences of population growth, will largely depend upon the institutional framework (Repetto and Holmes 1983; Cain and McNicoll 1988; McNicoll 1989; Ghimire 1993). It follows that assuming the variables P. T. and C' independent is far from neutral: it supposes a certain organization of society and of the world, typically one in which resources are freely and/or openly accessible. It is in such a context that Hardin wrote on the 'tragedy of the commons', although what he really meant by commons was open-access resources (Hardin 1968). In such a setting indeed the impact of the growing population will theoretically be at its maximum. Classic examples for such open-access resources are the atmosphere and the oceans. But paradoxically, it is with respect to these resources that the case of population growth as a main cause of environmental degradation is empirically the weakest. Extending this framework to resources like land or forest is very problematic because it does not take into account institutions that do exist, and the role of which is precisely to render interdependent the variables P. T and C'. The effect of land tenure, of community structures, of governmental intervention are not taken into account, although it is repeatedly argued that they play a crucial role. Indeed, in the static framework, Ehrlich's equation has the validity of an identity. But through the institutional framework, including the market system, the variables P. T and C' are made interdependent and the use of the equation to assess the contribution of population growth to environmental degradation becomes highly problematic.


The other goal of the Ehrlich equation is to guide environmental policies. By assessing the respective contributions of the different factors, it implicitly points to where adjustments are needed. Yet, when applied, the equation assesses the contribution of each factor during a limited period of time only, without taking into account the historical process that led to the environmental crisis. Considering the environmental impact at a given point in time, say 1960, it can only gauge the contribution of each factor at that time to a subsequent increase in this impact. This methodology is again very problematic. For instance, even if population growth had been the main factor leading to an increase in additional emissions of CO2 thereafter, the reason why this constitutes a problem is because the 1960 level of emission was already unsustainable, and this possibly for reasons quite independent of past population growth. The results obtained from applying the Ehrlich equation during a short time span cannot therefore guide directly future policies.

In this respect, an interesting distinction has been made by Shaw (1989). To contrast the intellectual roots of Commoner's position - that 'environmental impact is not correlated with the rate of population growth' (cited in Shaw) - with that of the other extreme and apparently opposite one of Nafis Sadik (1988, cited in Shaw) who claims that 'high fertility and population growth are contributing to the damaging of the natural resource base', Shaw introduces the distinction between 'ultimate' versus 'proximate' causes of environmental degradation. He then suggests that Commoner's claim is one of ultimate causality in which the major contributor to environmental degradation is the high and wasteful consumption of countries with low population growth rates, while Sadik's claim is one of proximate causality, where the question is not about the factors that produce a particular outcome, but those that aggravate, trigger or catalyse the incidence in a specific situation.

The conclusion Shaw draws from this distinction is quite typical. His argument is that Sadik's position has become particularly important today because the control of ultimate causes of environmental degradation have remained out of reach. In other words, between the two forms of
expansion of human activity, demographic and economic, Shaw sees the former as easier to tackle than the latter, and concludes that it should therefore be given priority. This view is not an isolated one. We find it, for instance, in the conclusion of Ehrlich et al(1993), where their discussion of the difficulties to meet future food demand is quite convincing. Yet they fail to address directly the issue of distribution, and only emphasize the population growth factor. As they conclude:

It is impossible to avoid the conclusion that the prudent course for humanity, facing the population-food-environment trap, must above all be to reduce humane fertility and halt population growth as soon as humanely possible (pp 24-5)... In theory much could be done to reduce the maldistribution of food, although doing so is certain to be very difficult in practice (p 26).

A last example is Bongaarts (1992) on the emissions of CO2: 'Since few governments are likely to adopt policies that deliberately reduce the growth in GDP per capita, any reduction in [growth of fossil fuel consumption] must preferably be brought about by reducing population growth or energy and carbon intensities' (pp 25-6).

The picture is now quite clear: under the assumption that ultimate causes of environmental degradation are out of reach, proximate causes like population growth are put at the forefront. The cynism of this view is striking: it calls for putting yet another limit on the poor who supposedly have too many children, whereas they already face limitations in all directions. It is thus those who have benefited least and suffered most from all the past experiences of development and of globalization who should make the necessary adjustments to the environmental crisis. Note that the assumption that 'nothing can be done about the ultimate causes' is in accordance with assuming that the variables P. T. and C' are independent.

But if this 'pragmatic' assumption is dropped, Shaw's distinction leads to a very different conclusion and points out another failure of usual applications of the Ehrlich equation. Ultimate causes are what created the environmental crisis in the first place. Those social groups or countries behind those causes should therefore bear the cost of finding and implementing a solution. Such a solution would be an institutional structure regulating the use of certain resources. To clarify the issue, we have to come back to the general causes of environmental degradation. The environmental crisis stems from the convergence of three factors: the finiteness of the natural resource base; a growing demand for natural resources; and inadequate institutions to regulate the use of these resources. Typically, the form of management of the resources should respond to the tension between a sustainable supply of resources and demand for these resources. And if demand is low compared to the resources available, there is no need for formal management.

Both the Ehrlich equation and the view that nothing can be done about the ultimate causes suppose that there is no regulating mechanism, and thus that all the blame for environmental degradation should be put on the growing demand. This absence of regulating mechanism is clearly a cause of environmental degradation, and it seems suprising to suppose that sustainability could be attained without doing something about it. It follows that the contribution of different factors to the environmental crisis cannot be gauged according to their direct impact on the environment. A different methodology must be used, one which distinguishes between two stages:

  1. Total demand is inferior to the sustainable threshold. In this case, there is no specific need to regulate the use of the resources. Different factors (population, consumption, technology) will affect total demand. To measure the contribution of each factor to an increase in total demand, the Ehrlich equation can be used with the proviso made in the preceding section.
  2. Total demand is above the sustainable threshold. If sustainability is held as a first priority on the political agenda, then the only cause of environmental degradation is the absence of an efficient regulating mechanism. That is, environmental degradation should be attributed entirely to a lack of political commitment to find an appropriate solution.The impact of such factors as population, consumption or technology should no more be assessed in terms of environmental degradation, but rather should be assessed with respect to the stress they put on the regulating mechanism, whether it exists or not.

The point is that the impact of population growth - or, for that matter, of increases in consumption per capita - cannot be assessed in the same way whether it takes place in an open world or whether it takes place in a closed one. For example, if A is the impact of population growth in the absence of an institutional framework, and B is its impact when there is an adequate institutional framework (A>B), B is the blame that should be attributed to population growth, A minus B being attributed to the absence of institutional framework. The polluter should be held responsible, not only for polluting, but also for creating the conditions in which others will pollute more if no solution is implemented. It does not make sense to denounce the unsustainability of population growth in societies which are unsustainable for altogether different reasons.

In this perspective, it makes little sense to speak about the relative contributions of population or consumption to increases in CO2 emissions in a country like the United States. The fundamental problem is that total demand has risen above the threshold under which no form of management is compatible with sustainability. Thus, the main cause of environmental degradation is that emissions of CO2 are not regulated.

Taking this approach, all the increase in emissions of CO2 in developed countries since 1960 would be due to a lack of political intervention, since in 1960 emissions per capita in those countries were already above the sustainable level with respect to the 1960 population. In these conditions, the contribution of population growth to the increase of world emissions would be 17 per cent (the share of developing countries, see (Table 5.3), whereas the blame for a lack of political intervention would be 62 per cent (total share of developed countries in the increase, see Table 5.1)).


A first goal of the Ehrlich equation was to clarify the controversy around the causes of environmental degradation. Is this goal achieved? At a purely numerical level, the index number problem leads to some ambiguities that should be clarified. At a more theoretical level, the equation leaves out a number of factors bearing on the environment; notably commercialization and the breakdown of traditional resource management systems. Finally, the heterogeneity problem leads to numerical results which are wholly misleading. This was very clearly illustrated by Table 5.3: we showed that population growth in Southern countries could not account for more than 17 per cent of the increase in CO2 emissions since 1960, instead of 64 per cent if the heterogeneity issue is not considered! And even 17 per cent is an overestimate since it does not take into account the problem of heterogeneity within the South. Speaking about population growth at the world level can only bring confusion, because it associates people with radically different life-styles. When we speak about population growth at the world level, it is obvious we speak first and foremost about population growth in the South. Yet when we calculate the contribution to the environmental crisis of population growth taking place in different regions of the world, the share of that in the North turns out to be greater than that in the South. There is therefore an important problem of semantics, which can only be solved through great cautiousness in the use of different terms. A first step towards clarifying the debate will be to fully acknowledge that it is meaningless to speak about the impact of population growth at the world level.

A second goal was to help design policies toward sustainability. The problem here is that the Ehrlich equation tends to blame the 'proximate' rather than the 'ultimata' causes of environmental degradation, to use Shaw's terminology. But this is again very misleading. The logic that says 'we cannot do anything about the ultimate cause, so let us do something about the proximate ones' supposes that there is a 'we'. It supposes the existence of a political community confronted with a series of different issues. By contrast, the global environmental crisis will require the creation of a global community. Certainly not a world government; but a community of states which recognize that they are addressing a common issue. Ever since the publication of the Brundiland Report entitled Our Common Future, it is more and more argued that we all share one world, and thus a common future. This, supposedly, creates some duties for all of us with respect to one another. It is also clearly a call for responsibility. Since we all share the same world, we all have to make some efforts to save it, and thus, implicitly, we all have to make some sacrifices for it. This world we share is a world of limited natural resources, of limited global commons. In other words, what we share is a common problem, the global environmental crisis, because it may affect the lives of us all. Furthermore, through this crisis, we all become supposedly interdependent: each person's behaviour and well-being is connected to other people's behaviours through the global biospheric system.

For the North, this sense of interdependence is perhaps quite new.

Certainly not so for the South who has felt dependent upon the North for many years, notably through economic relationships. What this means is that the 'ecological' interdependency revealed by the global environ mental crisis cannot be perceived in its own teens only. It must be recast in the history of international relations, and within the context of another form of interdependency, namely economic interdependency. In this perspective, addressing first the issue of economic relations appears a much more promising route to building some form of international cooperation, than speaking about population growth.


Binswanger, H and Ruttan, V (1978) Induced Innovation: Technology, Institutions and Development Johns Hopkins University Press, Baltimore

Bongaarts, J (1992) 'Population growth and global warming' Population and Development Review, vol 18, pp 299-319

Boserup, E (1965) The Conditions of Agricultural Growth George Allen and Unwin, London

Cain, M and McNicoll G (1988) 'Population growth and agrarian outcomes' in Lee et al (eds) Population, Food and Rural Development Clarendon Press, Oxford Coale, A and Hoover, E (1958) Population Growth and Economic Development in Low-Income Countries Princeton University Press, Princeton, New Jersey Commoner, B (1971) The Closing Circle Alfred A. Knopf New York

Commoner, B (1988) 'Rapid population growth and environmental stress' paper presented to the United Nations Expert Group on Consequences of Rapid Population Growth, United Nations, New York

Demeny, P (1986) 'Population and the invisible hand' Demography vol 23
Ehrlich P and Ehrlich A (1990) The Population Explosion Touchstone, New York

Ehrlich P. Ehrlich, A and Daily, G (1993)'Food security, population and environment' Population and Development Review, vol 19

Ehrlich P and Holdren, J (1971) 'Impact of population growth' Science, vol 171, pp
Ghimire, K (1993) Linkages Between Population, Environment and Development
UNRISD, Geneva

Hardin, G (1968) 'The tragedy of the commons' Science, vol 162, pp 1243-48
Harrison, P (1992) The Third Revolution: Environment, Population and a Sustainable
I B Tauris, London

Hayami, Y and Ruttan, V (1971) Agricultural Development: an International
Johns Hopkins University Press, Baltimore

Keyfitz, N (1991) 'Population and development within the ecosphere: one view of the literature' Population Index, vol 57, pp 5-22

Lappé, F and Collins, J (1980) Food First, the Myth of Scarcity Candor Book
Souvenir Press, London

Lewis, A (1955) The Theory of Economic Growth Allen and Unwin, London
Lorimer, F (1963) 'Issues of population policy' in P Hauser (ed) The Population
A Spectrum Book, Englewood Cliffs, New Jersey

McNicoll, G (1989) 'Social organization and ecological stability under demographic stress' in McNicoll and Cain (eds) Rural Development and Population: Institutions and Policy. The Population Council, Oxford University Press, New York

McNicoll, G Y and Cain M (eds) (1989) Rural Development and Population: Institutions and Policy The Population Council Oxford University Press, New York

Myers, N (1991) Population, Resources and the Environment: the Critical Challenges United Nations Population Fund, New York

Notenstein, F W (1945) 'Population: the long view' in T Schultz (ed) Food for the World Chicago University Press, Chicago

Pearce, D (1991) 'Population growth' in D Pearce, E Barbier, A Markandya, S Barrett S. K Turner, T Swanson (eds) Blueprint 2: Greening the World Economy Earthscan, London

Repetto, R and Holmes, T (1983) 'The role of population in resource depletion in developing countries' Population and Development Review, vol 9, pp 609-32

Rufin, J-C (1989) L'Empire et les Nouveaux Barbares Pluriel, Paris

Sadik, N (1988) 1988 State of the World Population Report: Safeguarding the Future United Nations Fund for Population, New York

Sen, A (1981) Poverty and Famines Clarendon Press, Oxford

Shaw, P (1989)'Rapid population growth and environmental degradation: ultimate versus proximate factors' Environmental Conservation, vol 16, pp 199-208

Simon, J (1981) The Ultimate Resource Martin Robertson, Oxford

Stamper, M (1977) Population and Planning in Developing Countries: a Review of Sixty Development Plans for the 1970s The Population Council, New York

United Nations (1991) World Population Prospects 1990 United Nations, New York

United Nations (1990) Result of Sixth Population Inquiry Among Governments United Nations, New York

UNEP (1990) Unep Profile Unep, Nairobi

WCED (1987) Our Common Future Oxford University Press, New York

Contents - Previous - Next