Population People

Posted August 2000

Population and deforestation

Alain Marcoux
Senior Officer
FAO Population Programme Service (SDWP)

This document is drawn from Population and the environment: a review and concepts for population programmes. Part III: Population and deforestation, by Alain Marcoux. FAO, June 2000.


This document presents facts and hypotheses about how population growth and migration are linked with forest change, then examines what the assessment of those linkages suggests regarding ways to improve forestry, population or development policies. It reviews population-forest linkages and stresses their diversity, noting that consequently the role of population dynamics in a local setting may vary from decisive to negligible. Similarly, the impact of forest degradation directly affects highly variable numbers of people. Finally, suggestions are presented regarding population-centred studies that might help improve forestry and rural development policy design.

1. Introduction

Deforestation is one of the major environmental issues, not only in directly affected countries and locations, but also from a global perspective. The degree of international attention to deforestation is commensurate with the role of forests in the global, national and local ecosystems. Forests provide a wide variety of highly valuable ecological, economic and social services, including: the conservation of biological diversity; carbon storage; soil and water conservation; provision of employment and enhanced livelihoods; enhancement of agricultural production systems; and improvement of urban and peri-urban living conditions (FAO, 1999: ix). Obviously, these services differ widely in nature and therefore tend to be valued in different manners by different societies and different social groups. While some services are immediately visible, others are of a long-term nature and take their full sense only in the perspective of intergenerational equity - a critical viewpoint in defining and assessing the sustainability of human development.

These services seem to be most at risk where they are most needed, i.e. in fragile ecosystems which characterize many poor countries and areas in the developing regions (FAO, 1999: 24-33); in particular:

The role of population change in deforestation and forest degradation has been much debated. This note attempts to summarize the state of knowledge as to how population dynamics (i.e. population growth and migration) is related to forest change. It then examines what the assessment of those linkages suggests regarding ways to improve forestry, population or development policies.

2. Global trends in the forest resources

Estimates of forest area and other parameters are provided by national forest inventory systems.1 FAO analyses and periodically consolidates these data, based on a common classification system. The latest global inventory thus published is the Forest Resources Assessment 1990 (FAO, 1995). Estimates were later made for 1995 (FAO, 1997, 1999) and the Forest Resources Assessment 2000 is under preparation.

The area of the world's forests has been estimated at 3 454 million hectares in 1995, including natural forests (about 97 percent) and forest plantations (about three percent); this represents about one-fourth of the land area of the Earth. A little more than half of this total area is located in developing countries.

Between 1980 and 1995 the extent of world's forests decreased by about 180 million hectares (Mha). This was the result of a net loss of 200 Mha in developing countries and a net increase of 20 Mha in developed countries. During 1990-1995, there was a decrease of 65 Mha in developing countries and an increase of 9 Mha in developed countries, for a net loss of some 56 Mha (an area the size of countries such as Benin, Cuba, Honduras or Liberia). Annex 1 shows the estimated change in forest cover by country and region during 1990-1995. The main facts seem to be:

Fortunately, the pace of deforestation seems to be slowing at the global level as well as in developing regions: the estimate of forest cover change in developing countries indicates an annual loss of 13.7 Mha between 1990-95, compared with 15.5 Mha between 1980-1990. We should have a sound idea as to whether this is a real trend once the comparable data set on deforestation expected from the Forest Resources Assessment 2000 becomes available.2

However, this does not mean that this trend will be observed in all countries, and for some of those the hypothetical implications of these trends are worrying. At constant rates of deforestation, the forest cover would be down to half of its 1995 extent around 2004 in Lebanon and Jamaica; around 2005-2010 in Afghanistan, the Comoros and the British Virgin Islands; around 2014-2018 in St Lucia, the Philippines, Haiti, El Salvador, Costa Rica and Sierra Leone; and around 2019-2024 in Pakistan, the Bahamas, Paraguay, Thailand, Nicaragua, Jordan and Malaysia.

Annex 1 also shows indicators of the availability of forest resources per person and per rural resident. The first indicator measures overall population pressure from all potential in-country users. The second refers to the population that is more likely to affect forests through direct use (consumption of fuelwood, land clearing for housing or agricultural purposes etc.) and can be dependent on the productivity of the forest stock for some of its needs in food and energy.

Indicators vary very much among countries, mostly because of wide differences in overall population density. Per person availability of forest resources is highest in Oceania (3.24 ha) and South America (2.74 ha) and lowest in Asia (0.15 ha). Values at the country level range from almost nil to tens of hectares - the latter in some South American locations and in Gabon. Resources per rural resident are highest in South America (12.16 ha) and Oceania (10.93 ha), and lowest in Asia (0.22 ha). Many countries have less than 0.1 ha per rural resident. Regions where urbanization is well-advanced (developed regions, Latin America, Near East) fare better on this indicator - but, of course, urban populations also exert some pressure on forest resources through food and other requirements.

3. Causes of deforestation and role of population factors

3.1 Population density and forest cover: statistical relationships

In the course of the FRA 1990 global assessment, a procedure was needed to adjust forest cover observations - which had different reference dates from one country to another - to the reporting dates, 1980 and 1990. For this purpose, statistical relationships between observed forest area change and ancillary variables, including population, were analysed for tropical areas (FAO, 1993).

Interestingly, the analysis demonstrated a significant relationship between forest cover and human population density, so much so that population density was selected as the independent variable of the algorithm developed to interpolate or project forest cover change. From the static point of view, an inverse correlation between population density and forest cover should not be a surprise, since when humans settle in a wooded area they need to clear it for building shelters and cultivate land. There cannot be a dense forest cover where there is high population density, and vice versa. In a dynamic perspective, however, things are less simple. Should a strong relationship be expected between the processes of population growth and deforestation?

Annex 2 presents a scattergram of population and forest change rates during 1990-1995. Population data from United Nations (1999) and forest data from FAO (1999) for 168 territorial entities have been used; entities where the forest cover was nil in 1990 have been excluded. No systematic relationship is apparent between the two rates; indeed, the linear correlation coefficient (-0.19) is negligible.

It is not difficult to find explanations for the independence of the deforestation rate from the population growth rate. One explanation has to do with the extent of forest cover at the beginning of the period considered. If the settlement area has little forest cover to start with, population may increase without reducing that cover, as long as settlement growth can be accommodated through densification and food production can be increased through agricultural intensification. Another explanation has to do with the fact that some of the human activities that may contribute to forest loss are driven by factors that have little or nothing to do with local population growth, as we shall now see. Thus a given population density can cause different degrees of 'demographic pressure', and by the same token an increase in density does not necessarily result in a proportional increase in pressure.

3.2 The diversity of population-forest linkages

Let us first consider population-forest dynamics in the setting of subsistence agriculture. Figure 1 below provides a schematic representation of a common pattern of activities and outcomes in such a setting.

Figure 1
Factors of deforestation in a subsistence agriculture setting

This figure highlights three important factors in deforestation, namely:

By contrast with subsistence settings, another category of needs for land arises from the requirements of commercial enterprises:

In seeking to assess the potential implications of population change for community and commercial forestry at the country level - and a fortiori at the local level - a close look at the local typology of population-forest linkages is required. In effect the mix of factors that impinge on forest futures varies considerably and therefore the relevance of population growth may range from very high to very low depending on local conditions. Figure 2 below illustrates, in one location, some of the vast differences that may exist in implications of human occupancy (Marcoux and Drigo, 1999).

Figure 2
A pioneer front area: Rondônia (Brazilian Amazon), 1996

That large forest tracts have been cleared is evident. Interestingly, however, three markedly different types of human impact are readily apparent:

Clearly, the statistical association found at high levels of aggregation between population density and forest cover embodies a heterogeneous array of linkages. Therefore, more detailed analyses are crucial when it comes to understanding and intervening on processes. Land tenure, local political institutions, and the market relations in which local producers are engaged would be relevant parameters in such analyses besides population characteristics and dynamics.

3.3 Role of population dynamics: growth and migration

As a rule, therefore, it is impossible to argue satisfactorily that population - or any one of the many factors that contribute to deforestation - is generally decisive. The diversity of situations is too great to support any generalization. A sensible warning against the temptation to ascribe simple blame in complex situations is that of Harrison (1992): having described a typical slash-and-burn, shifting agriculture system with rapid population growth, he concludes:

"It would be easy to come away … with [the] view that population growth was the sole cause of deforestation … But it would be an oversimplification. For if technology gradually improved yields in line with population growth, there would be no need to clear more forest to feed more people. It would be equally easy to come away convinced that technology lag, due to individual and national poverty, was to blame. These in turn are aggravated by declining commodity prices and an intolerable debt burden. Or one could blame the lag, the debt and the poverty on misguided government policy, which controlled markets and prices paid to farmers, and made it uneconomical to intensify. But these, too, would be oversimplifications. For if population were not growing, then there would be no need to clear new forest, even if no fertilizers or high-yielding varieties were used. Deforestation [takes place] because population growth is outstripping change in technology. The speed of both relative to each other is what counts. It is meaningless sophistry to claim that one is more fundamental than the other, or to call the one a 'root' cause and relegate the other to a mere 'exacerbating factor'."

3.4 Population and forest degradation: a regional view

Against this background, let us review the dominant forms of change in land use and their population connections, based on an empirical assessment of forest degradation in the major regions. The 1990 Forest Resources Assessment assessed land use changes on a sample basis, using remote sensing techniques (FAO, 1996). For each sampled location, the changes were assessed by interpreting two satellite images acquired around 1980 and 1990. Matrices describing class-to-class changes in land cover were established for a sample location, then standardized and aggregated, enabling assessments by geographic region and by ecological zone. Such information enables understanding complex processes that take place simultaneously, such as forest fragmentation, forest degradation in general (i.e. loss of biomass), afforestation etc.

Table 1
Pan-tropical area transition matrix 1980-1990

Table 1 synthesizes the changes thus assessed for the tropical areas as a whole. It shows that deforestation involves complex processes. The transition from a pristine forest cover to permanent agricultural use, implying total loss of that cover, usually is achieved through successive stages where the forest is increasingly depleted and the intensity of agricultural use grows gradually.

The losses of a given class to others are read along the related row of the table (e.g. closed forest: between 1980 and 1990, 92.05 Mha [1 367.96 minus 1 275.91] of the original closed forest cover were transferred to other classes; 8.97 Mha changed to open forest; 9,27 Mha to long fallow; etc.). The gains from other classes are read along the relevant column (e.g. the closed forest stock grew by 0.86 Mha transferred from open forest; 1.10 Mha from long fallow; etc.). As expected when considering such a diversified area, almost all the possible class-to-class changes are represented. However, most of the changes are located above the diagonal. Since the land cover classes are listed in order of decreasing biomass density, this pattern means forest degradation.

We can gain some empirical insight into the effects of population change by examining the most important types of land use changes revealed by the table:

With this background in mind, let us briefly review the regional patterns of land use change revealed by the above study and the role population pressure seems to play in those changes (FAO, 1996: 59).

4. Relevance for policies

4.1 Population policies and forest outcomes

One question is whether population policies and programmes can improve the outcomes of forest conservation and development policies. Let us look into this question from two angles, namely policies that address the population growth factor and those that address population distribution.

The value of standard population policies in the context of environmental issues is the subject of an animated debate. Some argue that stopping population growth is a necessary - albeit not sufficient - condition for environmental sustainability; others deny population pressure a major role in environmental degradation and cite inequalities in resource distribution, poverty, or the lack of employment opportunities.

However, it is important to realize that the policy value of these opposite stances do not depend only on the factual value of the explanation of environmental degradation upon which they are based. Proponents of the latter stance, for instance, suggest that environmental relief calls for attacking in priority lack of employment and poverty; but this is redundant, because those evils ought to be attacked in their own right. In addition, the role of poverty - versus affluence - as a factor in environmental degradation is highly debatable (Marcoux, 1998). They also suggest that attempts to influence population growth are pointless; but environmental conservation rarely if ever is an objective of fertility policies, which have strong health and social justifications. In any event, it is better to tackle all the causes of a problem rather than a few, especially when a vicious circle seems to exist (environmental degradation often increases the needs for family labour, providing an incentive for continued high fertility, etc.).

Could environmental conservation, then, be an additional motivation to undertake such policies? It would seem so, but an important caveat is necessary here: if one assumes that slower population growth will enable a faster growth in per caput income and consumption (and in fact this often is the main reason for wanting to slow down population growth), then logically one must doubt that a moderating effect on natural resources consumption will ensue.

At any rate, given that policies to moderate population growth are in place virtually in all developing countries, what effect could these policies have on deforestation? It would seem that the effect should be quite positive in situations where deforestation results from a combination of population pressure and stagnating economic, social and technological conditions. On the other hand the effect could be much weaker in cases where deforestation is driven by commercial ranching and extensive logging. Obviously, the magnitude of the effect of slowing population growth depends much on the nature of the mix of deforestation factors; one may distinguish three categories of factors in this respect:

Let us note that the population growth rate may not be the most critical variable. It is customary to compare that rate with the rate of GDP growth in order to assess progress in per caput income. This approach is not ideal in the context of natural resource policies, where one rather is concerned with physical quantities - say, with the ratio between the sustainable output of each critical stock of resources and the withdrawals that a growing population will operate from the said stock. In this case, it is absolute population increase that matters most. In addition, the GDP approach itself is not appropriate for sub-national levels, while the issue of deforestation is highly location-specific (and the need for limiting population growth may be clearest in specific rural settings where growth rates may be several times higher than the national rate).

One conceivable strategy to tackle the population growth factor is to apply special fertility control efforts in the area concerned. Another strategy, of course, is to encourage emigration from the area concerned, as this also results in a lower population growth rate. Seeing population and resources distribution from the country perspective, this strategy corresponds to the idea of equalizing population pressure across space.5 It is far better, however, to examine it from a local angle.

Attempting to divert population pressure and migratory flows away from forest areas seems all the more sensible as forests in all ecological zones have a lesser capacity per area unit to support a population sustainably than have croplands and grazing lands, despite the greater biomass carried. On the other hand one cannot help but remark that, on the contrary, government attempts to influence population distribution for rural development purposes often target forests as destination areas when undertaking to transfer populations out of densely settled zones (where, incidentally, these usually live from permanent agriculture rather than shifting cultivation). At any rate, the limited effectiveness of such attempts in terms of stability of the new settlements is a rather well established fact.

4.2 Population research and forestry policies

Long-term viability of the use of natural resources is a traditional concern among foresters, a community where the term sustainability actually seems to have appeared for the first time.6 How can population knowledge and expertise help ensure sustainable forestry development?

The UNCED recognized that population-oriented research can contribute to situation assessment and policy formulation on environmental issues. Agenda 21 provides useful ideas on the role of population factors in environmental policy making and therefore on their place in policy studies. This of course applies to deforestation as well as to other issues, the eventual aim being the formulation of integrated forest policies taking into account population concerns. Population-related work will be useful at all stages of policy making and implementation: assessment of forest development issues; policy design and programming; and programme monitoring and evaluation.

Human factors are important to identify priority areas for action. Besides technical factors and indicators on the physical condition of forests, certain population-related variables bring additional light on the issues at hand. In particular, in deciding on what value one will ascribe to the conservation and development of a forest area, the question of how many persons have a stake in the fate of that area (whether residents, direct users or indirect beneficiaries) is a legitimate consideration.

In this context some qualitative considerations are as important as population numbers, since the manner in which human populations could be affected by anticipated change in forest conditions must be a crucial consideration. In this spirit, Agenda 21 highlights the need "to assess human vulnerability in ecologically sensitive areas". Incidentally, it is worth reflecting on the distinction between vulnerable areas and vulnerable populations (meaning populations vulnerable to environmental damage): since local environmental damage very often has external consequences through ecological linkages (e.g. down water streams, down hill slopes, or along food chains), vulnerable population groups are not only those living in vulnerable areas but include those whose resources are indirectly affected by that damage.

Coming to agriculture-forest interactions, even at country level it will often be indispensable to consider the range of local agricultural systems with a view to assessing whether, and where, the growth in population density is likely to cause increases in demand for cultivable land for shifting or permanent exploitation, leading to more forest clearance. Finally, in a broader perspective, one may have to look beyond the populations living around or in forest areas and examine whether population-resources imbalances in more distant regions could not affect forest areas through migratory flows.

Issues regarding the sustainable development of natural resources often are best analysed and tackled at the local level, because the great variety of situations and causal factors precludes general solutions to environmental problems. This consideration has led in particular to the collaborative management approach, born out of experiences in rural development and natural resource management carried out during the 1980s by various international agencies and NGOs. Population issues initially were not a major concern in these experiences, but later it became clear that population growth and migration could not be ignored in attempting to promote sustainable human development at the local level (Warren, 1998).

An effort has indeed been made to provide methodological suggestions to practitioners for incorporating population issues in collaborative management processes. The outcome (Barton et al., 1997) describes the interactive nature of population change, environment and development and the collaborative management/'primary environmental care' approach; presents concepts related to population dynamics and the environment and introduces terminology and indicators; discusses issues of participation in development and describes general features, methods and techniques of participatory management and participatory action research; describes the phases of information gathering and appraisal of a comprehensive participatory action research on population dynamics and the environment, and introduces specific methods and tools; suggests simple methods and procedures allowing local actors to extract elements for planning from appraisal findings; examines issues related to the implementation and monitoring of activities and describes methods for evaluation enabling a 'learning-by-doing' process.

In general, 'population and development' programmes would do well to promote and assist with some of the above lines of work, promoting the required interdisciplinary research and emphasizing community-level experience wherever applicable. Again based on the Agenda 21 framework, three lines of research seem useful, namely:

In brief, one seeks a better basic information on forest use patterns and the impacts of forest degradation on human populations, and applications of that information in policy-making. A research strategy could for instance include the following types of studies and address the following questions:

In the longer run, accumulated knowledge and comparative studies will enable "a better understanding of the relationships among demographic dynamics, technology", culture and natural resources (United Nations, 1993).


1 Forests are defined as ecosystems with at least 10 percent crown cover of trees (or/and bamboo), generally associated with wild flora, fauna and natural soil conditions, and not subject to agricultural practices. Deforestation refers to change in land cover involving the depletion of tree crown cover to less than 10 percent.

2 The FAO global forest resources assessment 2000 will be the most comprehensive yet, "including information not only on forest cover but also on non-wood forest products, protected areas, plantations, volume and biomass, removals, ownership and fires, as well as studies on causes of forest change; an independent and objective survey of forest cover change based on remote sensing; and global maps of forest cover, ecological zones and protected areas" (United Nations, 2000).

3 See IUCN - FCP Forest Conservation Programme

4 For instance, in 19th century Europe, forests were more heavily affected by the development of coalmines and railways than by the rather slow population growth.

5 Besides a halt to population growth, proponents of the 'natural science perspective' recommend a "balanced population distribution", supposed to ensure an even pressure on natural resources. This idea is fraught with problems. In fact, reflecting about its possible criteria, methods and concrete policy implications rapidly reveals its superficiality (Marcoux, 1998).

6 In Sylvicultura Oeconomica, a 1713 book by Hans von Carlowitz, Director of the Saxonian State Forests.


Annex 1
Forest cover indices (1990-1995) by country and region

Annex 2
Population and forest change rates, 1990-1995


Barton, Thomas, Grazia Borrini-Feyerabend, Alex de Sherbinin and Patrizio Warren. 1997. Our people, our resources. Supporting rural communities in participatory action research on population dynamics and the local environment (pdf file, 54K). Gland, IUCN-UNFPA. (Spanish)

FAO. 1989. Forests, trees and people. Forestry Topics Report No. 2. Rome.

FAO. 1991. Sustainable agriculture and rural development in Asia and the Pacific. Regional document No. 2. FAO-Netherlands Conference on Agriculture and the Environment, 's-Hertogenbosch, 15-19 April 1991.

FAO. 1993. Forest resources assessment 1990. Tropical countries. FAO Forestry Paper 112. Rome.

FAO. 1995. Forest resources assessment 1990. Global synthesis. FAO Forestry Paper 124. Rome.

FAO. 1996. Forest resources assessment 1990. Survey of tropical forest cover and study of change processes. FAO Forestry Paper 130. Rome.

FAO. 1997. State of the world's forests 1997.Rome.

FAO. 1999. State of the world's forests 1999. Rome.

Harrison, Paul. 1992. The third revolution. Environment, population and a sustainable world. London, Tauris & Co.

Marcoux, Alain. 1998. From linkages to policy issues: population and environmental change. Paper presented at the Thematic Workshop "Population, poverty and environment", Rome, 26-30 October 1998. FAO.

Marcoux, Alain and Rudi Drigo. 1999. Population dynamics and the assessment of land use changes and deforestation. Rome, FAO.

UNEP. 1992. The world environment 1972-1992. Two decades of challenge (M.K. Tolba et al., eds.). London, Chapman & Hall.

United Nations. 1993. Agenda 21. The United Nations programme of action from Rio. New York.

United Nations. 1997. World urbanization prospects. The 1996 revision. New York.

United Nations. 1999. World population prospects. The 1998 revision. New York.

United Nations. 2000. Integrated planning and management of land resources. Report of the Secretary-General. Addendum: Combating deforestation. Economic and Social Council (E/CN.17/2000/6/Add.1).

Warren, Patrizio. 1998. Collaborative management of human development, natural resources and population. Paper presented at the Thematic Workshop on Population, Poverty and Environment, FAO-UNFPA, Rome, 26-30 October 1998.

World Energy Council and FAO. 1999. The challenge of rural energy poverty in developing countries. London, World Energy Council.

Back to Top FAO Homepage