|
THE POLICY AND ECONOMICS OF LAND USE IN THE BOLIVIAN LOWLANDS:
A CASE STUDY OF FORESTRY AND CATTLE.
Frank Merry
School of Forest Resources
and Conservation
University of Florida
Gainesville, Florida
April 25, 2001
SECTION I
BACKGROUND RESEARCH AND DISCUSSION
SECTION II
PAPERS
1. Accounting for regional productivity, forest management costs, and land use taxes in the
analysis of land conversion from forest to pasture in the Bolivian Lowlands.
2. Can increasing net returns per hectare in sustainable forestry exacerbate conversion to
pasture? A case study in the Bolivian Lowlands.
CONTENTS: SECTION I
Introduction to Eastern Bolivia 1
Economic concepts in land-use 1
Land markets and prices 6
The economics of tropical forestry 8
The economics of deforestation 12
Forest institutions 13
Government policy and forest management 14
Legislation affecting land use in Bolivia 18
Agricultural legislation 19
Forestry legislation 23
Further reforms in land taxation 30
Land use planning requirements 32
The cattle and forestry conflict in Latin America with implication for Bolivia 33
Cattle management in the Bolivian Lowlands 35
Beef markets and prices 38
Production activities, timetable, and budget 39
Land clearing for pasture 41
Grasses 44
Cattle and beef transportation 46
The forestry sector: description and issues 48
Export markets 49
Domestic markets 52
Timber prices 53
Forest management costs 53
Management plans, inventories, and operating plans 54
Timber transportation 56
Eastern Bolivia
Population growth, the discovery and mining of mineral resources, improvements in the
transportation system, timber extraction, and the emergence of an industrial agricultural sector have
all contributed to the extraordinary change, in the Eastern Lowlands of Bolivia, a transformation
which is likely to continue into the foreseeable future. Notable in this change has been the explosion
of agricultural and forestry activity in the Santa Cruz. From virtually nonexistent, current
agricultural and forestry production now comprise 2 percent of Gross National Product, and
approximately 20 percent of Gross Departmental Product for Santa Cruz. For a nation whose
modem economic history had been based on mineral extraction in the highlands, these figures
represent an important diversification of the economic base.
Agriculture and forestry are now prominent factors in determining land use patterns in Santa
Cruz. Of the total 36.5 million hectares that constitute the Department of Santa Cruz 5.4 million
hectares has agricultural potential with varying degrees of production limitations and an additional
2.6 million has limited agricultural potential. Approximately twenty-eight million hectares are
suitable for either beef production and/or forestry. With only 750,000 hectares currently under
agricultural production, and five to six million allocated to forest concessions, it would appear that
there is capacity for growth in both agricultural and forestry sectors. Assuming that economic
development is a desired goal for the region, and to ensure that the land resource is at once
productive and enduring, a balance between development and resource conservation must be
realized.
Introduction to economic concepts in land-use
1
Eastern Bolivia
Population growth, the discovery and mining of mineral resources, improvements in the
transportation system, timber extraction, and the emergence of an industrial agricultural sector have
all contributed to the extraordinary change, in the Eastern Lowlands of Bolivia, a transformation
which is likely to continue into the foreseeable future. Notable in this change has been the explosion
of agricultural and forestry activity in the Santa Cruz. From virtually nonexistent, current
agricultural and forestry production now comprise 2 percent of Gross National Product, and
approximately 20 percent of Gross Departmental Product for Santa Cruz. For a nation whose
modem economic history had been based on mineral extraction in the highlands, these figures
represent an important diversification of the economic base.
Agriculture and forestry are now prominent factors in determining land use patterns in Santa
Cruz. Of the total 36.5 million hectares that constitute the Department of Santa Cruz 5.4 million
hectares has agricultural potential with varying degrees of production limitations and an additional
2.6 million has limited agricultural potential. Approximately twenty-eight million hectares are
suitable for either beef production and/or forestry. With only 750,000 hectares currently under
agricultural production, and five to six million allocated to forest concessions, it would appear that
there is capacity for growth in both agricultural and forestry sectors. Assuming that economic
development is a desired goal for the region, and to ensure that the land resource is at once
productive and enduring, a balance between development and resource conservation must be
realized.
Introduction to economic concepts in land-use
1
Economics is defined as the study of the allocation of scarce resources to best accommodate
human wants and needs. As a subset of that field, land economics deals with the allocation of a
particular resource land and how this resource is best used to provide economic benefit. This
section presents some of the fundamental concepts of land economics theory, the most important of
which is land rent.
Land rent is defined as the economic returns that accrue to land for its use in production, or
the residual economic surplus that is available after accounting for total costs of production (Barlowe
1972). Economic theory suggests that a firm should continue to produce until marginal costs equal
marginal revenue for any production unit. Rent is gained when marginal costs are less than marginal
revenue. If the ability to produce at a marginal cost less than marginal revenue is attributable to the
land then it can be said that land rent has been earned.
Land rent provides the theoretical basis for the value placed on land and real estate resources.
It can also provide incentive for ownership, influence allocation of land between competing uses,
and affect leasing, taxation, and development policies. In addition, when choosing between two land
use alternatives, e.g., agriculture or forestry, that which provides the highest land rent will be the
preferred option. Usually attributed to the British economist David Ricardo ':
If all land had the same properties, if it were unlimited in quantity, and uniform in quality,
no charge could be made for its use, unless where it possessed peculiar advantages of
situation. It is only, then, because land is not unlimited in quantity and uniform in quality,
S The Principles of Political Economy and Taxation, 1817.
2
and because, in the progress of population, land of an inferior quality, or less advantageously
situated, is called into cultivation, that rent is ever paid for the use of it. When, in the
progress of society, land of the second degree of fertility is taken into cultivation, rent
immediately commences on that of first quality, and the amount of that rent will depend on
the difference in quality of these two portions of land." 2
Rent can be affected, however, through government legislation on taxes and zoning, and
infrastructure development; whereas Ricardo thought of land in terms of quality alone, it was the
economist Von Thunen who first described the described the importance of product characteristics
and location in land rent
Von Thunen's premise was that land would be allocated based on the characteristics of the
products. He assumed that all land was homogeneous and thus avoided the issue of land quality that
Ricardo so assiduously embraced. Von Thunen introduced the concept of concentric rings of
production. He states that if land surrounding the village is homogenous, production will be in rings
depending on the characteristics of the product. He suggests, for example, that intensive production
perishable crops would be placed immediately surrounding the village; the second ring would
accommodate the bulky items and those difficult to transport; and more distant rings could be filled
with items that were easily transported and had lower per unit value.
Additions to this line of reasoning can be made by considering changes in infrastructure and
transportation. Included within Von Thunen's assumption of homogeneity is infrastructure, i.e.,
2 David Ricardo, cited in Barlowe pg (1972).
3
transportation costs were the same for all regions. Imagine, however, the impact of a new highway
on transportation costs. Land closer to the highway will secure an immediate decline in
transportation costs which will, in turn reduce marginal costs of production and provide a
comparative advantage of production based solely on the introduction of new infrastructure. This
leads to the point or line where one product is favored over another referred to as the margin.
The concept of the margin is one of great importance in land use. Whereas in production
theory the marginal concept has to do with the addition of a unit of production, in land economics
the margin is essentially a boundary or frontier with different optimal land use on either side. On one
side of the margin it is better (more profitable) to produce a given product, and on the other side of
the margin, another. Again, the reasons why one product differs another has to do with land quality,
transportation, and the characteristics of the product itself; thus margins are determined by a
production rent gradient in which the most profitable land will be brought into production.
Successively poorer land will then be brought into production until the supply of productive land -
that land which produces rent is ended.
The supply of land can be considered from two perspectives, physical and economic. The
physical supply of land is, naturally, the land area on the planet, whereas the economic supply is that
land available for use by mankind. Because the physical supply land is unlikely to change, it is
prudent to concentrate on what shifts the economic supply of land. First one can identify several
natural limits to supply which include for example, sunlight, temperature, precipitation, topography,
and soils. Secondly, technology can affect land supply by altering the economic arena in which the
land based. The institution of government also plays a major role in determining land supply since
legislation can directly affect supply through such actions as zoning, ordinances, and taxation.
4
Land markets and prices
Land value, as mentioned above, is based on the willingness of people to pay for the land.
This willingness is in turn a factor of the productive potential, or expected land rent. If land is
bought and sold, and no productive activity takes place during that time, no moral code has been
broken. Rather, it is simply a matter of two different levels of willingness to pay, with the second
higher than the first. Land speculation cannot exist in a vacuum. The land must be thought valuable
based on its potential production.
The more open the market, the easier the transfer, and the more efficient the use of the
resource. Barlowe (1972) identifies several characteristics of the real estate market that drive how
this market functions:
* Limited product availability,
* Fixed location,
* Nonstandard products,
* Legal transfer requirements,
* Dependence on local supply and demand conditions,
* Large considerations with most transactions,
* Infrequent participation by many buyers and sellers,
* Use of credit to enable purchase,
* Common use of a broker.
In addition to its productive capabilities, land has traditionally been a good that people find desirable
6
because of its expected durability, long life and relative ease of management, all of which combine
to ensure that land has historically been a popular investment.
Poor infrastructure and low population combined to make certain that land has been almost
a free good in Eastern Bolivia. Times have changed, agricultural production is booming in Santa
Cruz, daily improvements in the infrastructure provide better access to markets, and land is no longer
a free good. In fact land is becoming scarce, and where previously it was possible to clear and claim
land as needed, one must now purchase through an active, but unregulated, market.
With poor records and informal land transfer, there is only vague, oral, evidence of land
values. Estimated prices vary from one to two thousand dollars for cleared land in the intensive
production zone Zonaa de production integrada) and the expansion zone Zonaa de expansion) around
Santa Cruz City. Prices decrease dramatically outside these zones, and drop to as low as $US 12 for
uncleared land toward the Brazilian border. In the area designated by the PLUS as for combined
cattle and forest production in the Chiquitano region, cleared and grassed land is sold at
approximately $US500 per hectare. Uncleared land in this region varies considerably but can be
purchased for as low as $US20 per hectare. In general it costs approximately $US 300-400 to clear
a hectare of forest for either arable production or improved grass (including fencing in grassing).
As mentioned above, these prices are largely undocumented in a land market that has received little
formal research.
Early government policies in Bolivia have tried to encourage development through migration
of indigenous people from the Altiplano to the eastern lowlands. The incentive was a large tract of
productive land available for those who chose to resettle. Initially, response to this incentive was
poor due to the lack of transportation which made trade outside the region uneconomical. Slowly,
7
Finally, the supply of land is strongly affected by infrastructure which directly affects the physical
location of land with respect to markets and transportation facilities and thus is fundamental in
determining land supply. The combination of these factors mentioned above produces a margin at
which land is productive. Beyond this margin, there will be no apparent demand for land as it is not
profitable to bring further land into production.
The demand for land is primarily a factor of population in conjunction with per capital levels
of food consumption and energy resource requirements. Simply put, as population increases and
living standards improve, the demand for land increases. Although Bolivia has a low population
density of 7.2 people per square kilometer, and a growth rate of 2 percent per annum, Santa Cruz
itself is in a different situation. The population of Santa Cruz City has gone from 279,000 in 1978
to 700,000 in 1992. This rapid growth of the urban population is accompanied by an increased
demand for agricultural and forest production, and, when coupled with access to external markets,
provides a root of land use changes. With an increased demand for land and with land area
somewhat fixed there is always competition between uses. However, it is assumed that land will go
to the highest and best use, which, all else being equal, is the land use that returns the highest rent.
Land, however, is merely an input to production and does not produce in isolation.
The three factors of production are land, labor, and capital. Land is therefore a factor of
production and will be productive only when the other two factors are applied. The amounts of labor
and capital applied varies in each production system, and are determined by relative returns. For
example if labor is cheap, there will be a greater propensity toward its use rather than substitution
for capital.
but surely, this situation has been altered to the present picture, where the primary recipients of land
are middle to upper income farmers/businessmen. With vast areas available, and an ease of moving
from one parcel to another, the value of land has been kept very low. Coupled with extremely poor
land titling estimates suggest that has much as 90 percent of agricultural land in Santa Cruz has
improper title there has been little or no incentive for land management.
Under the new ley INRA, there is a massive push to correct titles (saneamiento) throughout
Bolivia. It is estimated that upwards of 90 percent of productive land in Eastern Bolivia is without
proper title. User rights and informal demarcation are the common forms of ownership. Titles are
difficult, if not impossible, to obtain and even when properly documented there are cases of squatting
and falsification of ownership documentation. Additionally, the high costs of obtaining proper title
can be prohibitive for most land users.
The economics of tropical forestry
This section presents a preliminary overview of the economics of tropical forestry, broken
into subheadings of: economic analysis of tropical forestry; the economics of deforestation; forest
institutions; and the effect government policy on forest management.
Land in forestry has always demanded special attention in economic analysis. Several
features distinguish the economics of forestry (Klemperer 1996) the first of which time and while
forestry does not have a monopoly on the problem of time in economics, it certainly is a more
important issue in forestry than in many agricultural crops. With time such an important facet of
production, uncertainty plays a critical role in forestry decision-making. A simple comparison can
be made here between the annual returns from an agricultural crop, corn, and the possible rotation
8
of, say, mahogany, which can be anywhere from 60 to 100 years.
In temperate forest several species have rotation lengths similar to tropical hardwoods and
the same problem is apparent. The difference however, is that temperate species are planted in
monoculture stands thus have a very high density per hectare whereas tropical hardwoods are
found scattered throughout the forest with a consequently low density per hectare. The second
factor which distinguishes forest economics is the concept of the factory and the final product -
where the timber trees are both. After harvest the ability to produce is lost, except of course, with
coppice species. Again this is not unique to forestry, many products are held for expected gain, for
example, land. Other distinguishing features of forestry are negative externalities associated with
harvest; simultaneous production of many outputs, many of which have no market value; and
societal conflicts in forest management (Klemperer 1996) all of which apply to the management
tropical forests.
There has been much debate on the economic analysis of tropical forest management (Leslie
1987, Barbier et al 1991, Sandler 1993, Toman 1994). The Food and Agricultural Organization of
the United Nations (FAO 1985) outline broad conditions for natural management as: being
ecologically and technologically possible; economically feasible and attractive; and socially and
politically practicable. These conditions are at best difficult to identify and even more difficult to
achieve. The conditions under which the expected returns from natural forest management will
cover its costs have been identified as: when growth rates and stumpage values are at a maximum;
silvicultural costs are at a minimum; and the discount rate is well below that currently applied.
(Leslie 1987, Vincent 1990). Naturally this begs the question when can these criteria be met.
In fact, it may be difficult to meet most if not all of these criteria. The FAO outlines given
9
above are broad statements that merit no further examination. The conditions required for natural
forest management to cover its costs do merit examination simply because they are difficult, if at
all possible, to achieve. First growth rates and stumpage values by definition growth rates of an
old growth forest are zero. The primary tropical forest ecosystem has reached its zenith and is for
all intents a closed system. For growth rates to increase there must be a harvest. This will remove
all the mature individuals from the final growth stage before senescence. In a s-shaped growth curve
for the entire stand, the harvest action will move the forest away from the flat top of the curve
backwards to the steeper, faster growing, region. So, by removing the old trees and changing the
status of the forest from an "old-growth or primary" to a "productive" forest we have increased the
growth rates. It is believed that annual growth rates in tropical forest are low a maximum 5 percent
and in fact more likely to be around two or three. In addition, with few genetic improvement
programs and poor management it is unlikely that these rates will increase.
Land in forestry is valued by the potential earning, which in a natural forest management
program, is the annual income from available timber defined by the price minus the cost of
extraction of the estimated 2 to 5 percent of timber volume growth of an infinite number of years.
Due to low prices for many tropical timbers and high costs of extraction these stumpage values can
be low if not zero. Vincent (1994) argues that tropical timber prices are low because the scarcity
signals associated with resource depletion do not register for tropical timbers. He states that because
the timber markets are worldwide, and there is currently no shortage of timber, the overall price for
timber is kept low. And, although tropical timber may become more scarce, the signal that this
would normally bring to the marketplace a price increase is not executed. In addition, many of
the tropical hardwoods compete in markets replete with low-cost alternatives, such as flooring. Thus,
10
stumpage values are subject to timber prices, discount rates and silviculture costs.
The discount rate is the interest rate at which future values are discounted to present values.
It is determined by a mix of time preference and the opportunity cost of capital, which can be
expressed by current interest rates. The higher the discount rate the more less valuable future
income. Exact discount rates are often difficult to establish, and for that reason, in economic
analyses, are usually presented in ranges from 2 to 20 percent, and any variation therein. Any
discount rate below 5 percent is usually considered to a "social" discount rate and is used in an
attempt to capture the nonmarket value associated with environmental goods. More often, however,
personal discount rates are much higher.
In a country such as Bolivia, where uncertainty is rife and the interest rates are relatively high
at about 18-20 percent, it would be rare to consider discount rates below 10 percent. Therefore the
low discount rates mentioned by Vincent will unlikely be found in the developing world. An
interesting caveat here is that if a sustainable management program is being undertaken in the forest,
the discount rate plays less of a role. This is because on a sustainable yield program, the income is
annual the value of the 2-5 percent growth mentioned above. Thus the operator need not wait
twenty years for income, rather if the timber is as yet untouched, he will receive the initial flush from
the old-growth harvest and then the annual income from the value growth.
Finally, stumpage values can be increased by lowering silviculture costs. Many proponents
of reduced impact logging (RIL) techniques claim the improved management is more economically
efficient than the conventional logging system (Boltz 1999). It remains safe to say that RIL need be
considered on a case by case basis. It has also been suggested that many adopters of RIL may take
only those parts of the overall management plan that will increase short term profits, while avoiding
11
longer term costs.
The economics ofdeforestation
An issue apart from the management of tropical forests is that of deforestation. Deforestation
is essentially the changing from one land use to another, in this case a preference for a land-use
activity other than forestry. In most of Latin America the two culprits associated with deforestation
are pasture for cattle ranching and slash-and-burn activities for small farming activities. In Bolivia,
however, most of the deforestation is directly attributable to large-scale, industrial, agriculture.
The causes of deforestation are often broadly listed as a combination of factors such as of
population, agriculture, infrastructure, tenure, and government policies (Southgate et al. 1991).
Indeed, there is a growing concern that government policies are important contributors to the
mismanagement of tropical forests and deforestation; a debate centered around the issues of rent
capture and extra-sectoral incentives unsympathetic to long term forest management goals (Repetto
1987, Gillis and Repetto 1989, Vincent 1990, Barbier et al 1991, Binswanger 1991, Larson and
Bromley 1991, Hyde and Sedjo 1992, Gray 1992, Deacon 1994). In a study of 150 economic
models of deforestation, Kaimowitz and Angelsen (1998) conclude that deforestation is increased
by: an improved infrastructure; high agricultural and timber prices; low rural wages; and access to
international trade. In addition they submit that policies associated with economic liberalization will
generally increase the pressure on tropical forests. They further state that there is no consensus on
the impacts of tenure, land markets, population, and household income.
From a land rent perspective there are two important points. The first is the contention that
the change from forestry to agriculture is a result of high-grading the forest, which reduces the forest
12
value and thus encourages alternative land uses. The second is that increasing returns for agricultural
production, usually subsidized in one form or another, surmount the forest value and encourage
change.
Forest institutions
Forest institutions, ostensibly organizations for the improvement of forest management, can
indeed pose one of the more substantial barriers to better management. Managerial failures in public
administration, dependence on external finance, poor budgetary discipline, and lack of institutional
reform, were all cited as major causes of barriers to growth in forest production (Kelvin 1993). Thus,
the ubiquitous problem of poor forest institutions is one of deep concern.
Morell and Anziani (1994) describe forestry institutional development in three distinct
phases: first occurring during the first half of the century was based on the development of industry
through the appropriation of the value of the forest resource without concern for the conservation.
The second phase was from the mid-70's until 1985 during which time the value of the forest
resource as a tool for national; development was minimized. The impact on forest institutions was
a lack of clear direction and in many cases funding decreases. The third phase is described as one
with increasing environmental awareness and increasing on non-governmental organizations. Forest
institutions are undergoing restructuring and remain without a clear sense of direction and purpose.
Forest institutions in many developing nations are underfunded, inefficient, and without clear
direction. The FAO (1993) portrays a list of forestry related institutional constraints some of which
are given below:
* Lack of coordination and facilitation with local forest users;
* Lack of quality and interdisciplinary training;
* Lack of specialized career streams such as social forestry;
* Orientation to only timber production;
* Lack of linkages to a private sector;
* Inability to manage tenure and rent issues;
* Staff incentives poorly designed;
* Complicated planning systems which exclude participation of forest users, NGOs and local
communities; and
* Funding constraints.
Government policy and forest management
One of the concerns surrounding the effectiveness of government policy in managing the
topical forest concession is that of rent capture. Rent, defined here as the value in excess of costs of
production including normal profit, is the primary issue of discussion. Equivalent to "stumpage
value" in forestry operations, the rent from the tropical forest resource is, in many cases, being
captured by the holders of large timber concessions, creating above-normal profit, and in turn an
incentive to enter the industry and extract the resource. In poor or developing nations with large
forest resources the attraction to harvest that resource is inevitable. However, the concern is not that
part of the forest resource is harvested but that the returns from that resource are not captured by the
government.
In his opening paragraph, Gillis (1988:43) states that "Indonesian deforestation would have
14
been less rapid had government policies had more neutral effects on tropical land use decisions..."
Gillis separates government policies affecting the forest resource into two sections, those of forest
sector policies and non-forest sector policies. The forest sector policies include stumpage fees or
royalties which are assessed at a uniform ad valorem rate of 6 percent of posted prices for all logs
regardless of species; log export taxes aimed at improving domestic value-added production but
which in fact cost the government a considerable amount of money due to lost revenue in taxes and
lack of conversion efficiency in domestic mills; and concession and reforestation taxes which are
to low and not collected respectively. Non-forest policies include such efforts to improve the rest
of the economy as resettlement, general tax and agro-conversion subsidies, fuel subsidy, and
exchange rate policies. Gillis acknowledged that a differentiated royalty which assesses individual
levels of taxes for each species is an option, but recommends an ad valorem uniform royalty tax
because of its ease in collection and because he believes that there will be little difference in the
extent of high-grading conducted under both schemes. The problem remains the institutional ability
to monitor and collect either tax.
Vincent (1990) essentially agrees with the findings of Repetto and Gillis (1988) who state
that current royalty systems drastically undervalue tropical timber. He states that a royalty system
is economically efficient only if it captures the full value of the resource rent (1990:214) and
demonstrates that this is only possible if there is a differentiated royalty for each species such that
the royalty for each species is equal to the stumpage value. The site stumpage value is assessed by
summing the individual stumpage values. The price for timber is assumed to be constant and
uniform across species, and costs are divided among marginal and fixed. This implies that the
difference in stumpage values are due solely to differences in extraction costs. Vincent continues to
15
show how the rent captured by an undifferentiated royalty will be divided among government
revenues, windfall profits for the operator, and high grading. The lost value in high grading comes
through a loss in potential rent from individual trees due to a decrease in harvest volume.
Hyde and Sedjo (1992) disagree with the approaches taken by Repetto, Gillis, Vincent and
others stating, that their discussions confuse measures of short and long-run economic efficiency,
overlook lump-sum charges, administrative costs and forestry agency experience. Going even
further, Toman (1994) notes that in the discussions of sustainable forest management economists
should be careful to distinguish between efficient allocation of resources and socially optimal
allocations reflecting inter-generational equity. Hyde and Sedjo also state that the critical issues
affecting the available level of economic rents and possible means of collection in tropical forest
management are the levels of privately efficient harvest, the level of socially efficient harvest, and
the costs of governmental institutions to monitor the use of the resource.
Hyde and Sedjo then address the issue which type of royalty would be best suited to capture
the potential rents from the resource. The three options that much of the previous literature describes
are the fully differentiated royalty, the flat rate or ad valorem royalty, and the uniform fixed royalty.
The fully differentiated royalty distinguishes among tree species, tree quality, land quality and access
class and is intended to capture all possible rent between the market price and the marginal cost
curve. The unavoidable problem associated with this form of royalty is the level of information and
institutional support required for it to be effective.
The flat rate or ad valorem royalty are percentage charges on net revenues. There is no
change in the level or quality of harvest rather the returns to the operator are simply decreased by
the percentage of the royalty. Uniform fixed royalties are flat fees per unit of harvest (also area
16
fees). These may not be effective because Hyde and Sedjo argue that if the elasticity of harvest
volume with respect to the uniform fixed royalty is greater than one then increasing the fixed royalty
will in fact decrease government revenues. Paris and Ruzika (1993) suggest that environmental
taxes only encourage the short-term private profit maximization and do not encourage or reward
long-term sustainable management practices. This argument relies on concessionaires voluntary
environmental management which is unlikely unless the rewards are greater than the loss in revenue
due to the decrease in extraction.
The recommendation given by Hyde and Sedjo for capturing the rent, including institutional
and environmental costs are with lump-sum bidding, where concessionaires bid competitively for
the right to harvest the resource, or a combination of fixed royalty and competitive bidding. These
would still not preclude the need for information to be made available to the forest ministry. In a
paper on competitive bidding schemes in the Philippines, Paris et al (1994), addresses a scheme
currently under use. Described as forest performance bonds, which essentially lease out the forest
to the highest bidder but with stipulations on environmental management. In addition, there remains
a 25 percent ad valorem tax on the local FOB market price of the logs and other products.
All of the royalty forms mentioned above are not without merit. It is, however, important
to consider the practical side of forest concession management under the normal budget constraints
of a forest ministry in the developing world. And it is here that the simpler forms, such as the area
and the ad valorem tax, shine.
Legislation affecting land use in Bolivia
This section reviews the recent legislative changes affecting land use in Eastern Bolivia.
17
Primarily the new agriculture and forest laws passed in 1997.
Agricultural legislation
The revolutionary political changes of 1952 in Bolivia brought massive reform of agricultural
land tenure in the highland and valley regions. Land held in haciendas was broken into small parcels
and distributed to peasant farmers, transforming the land-use pattern to an agricultural sector based
on minifundia. Land reform had little relevance, however, in the lowlands of Bolivia. At the height
of the agrarian reform, virtually the only area in the tropical lowlands with any appreciable
population was the immediate agricultural area just north of the city of Santa Cruz. Land was not
scarce in the lowlands, but merely inaccessible. Even the significant changes following the
completion of the road from Santa Cruz to Cochabamba in 1956 were limited to the area surrounding
Santa Cruz.
In the 1960s, government policies encouraged migration from the Highlands and Valley
regions to the Eastern Lowlands. The incentives were large tracts of productive land available for
those who chose to resettle. Initially, response was poor, mostly due to the lack of infrastructure
which made trade outside the region uneconomical. But gradually both planned and spontaneous
colonization programs have taken hold. Besides the eastern lowlands of Santa Cruz, similar
colonization efforts reached from La Paz north into the lowland region called the Altobeni and from
Cochabamba north into the tropical lowland region known as the Chapare. Parallel to the
colonization programs in Santa Cruz, the agricultural "frontier" expanded to large proportions with
growth of medium and large scale farming enterprises on an area of perhaps a million hectares. As
land has taken on market values, tenure has only now become conflictive within and among groups
18
such as subsistence farmers, medium and large scale farming and livestock operations, logging
interests, and indigenous groups.3
With no clear geographical references on titles, no cadastral system, and a title registry
system indexed by owner rather than by physical land parcel, the inevitable result has become a
confusion of overlapping claims based on conflicting documents, often without regard to local
occupancy. It is estimated that almost 95 percent of agricultural land in Santa Cruz, for instance, has
improper title. The new agricultural law of 1997 was aimed to change all of this.
The Ley de Servicio National de Reforma Agraria (Ley No. 1715 known as the ley INRA)
has three principal aims. First is to establish the organizational structure and attributes of the
National Service for Agrarian Reform (SNRA) and the regimen of land distribution; secondly, to
guarantee land rights; and third, to create the Superinterdencia Agraria, the Judiciara Agraria
(Agrarian Judiciary) and detail their processes in order to regulate the titling process for agrarian
lands. The ley INRA comprises six major components or titles as follows:
Title one. National Service for Agrarian Reform
General Disposition
Title two. Institutional Format and Duties
Of the National Service for Agrarian Reform
Of the President of the Republic
Of the Ministry of Sustainable Development and Environment
3 Mining and petroleum rights continue to be administered through public concessions, as mineral rights
belong exclusively to the state, with no benefit accruing to the landowner.
19
Of the National Agrarian Commission
Of the National Institute of Agrarian Reform (INRA)
Of the Superintendencia Agraria
Of the Agrarian Judiciary
Constitution
Of the National Agrarian Tribunal
Of the Agrarian Judges
Title three. Agricultural Holdings and Land Distribution
Agricultural Property
Land Distribution
Title four. Reversion and Expropriation of Land
Reversion of Land (to the State)
Expropriation
Title five. Land Titles for Agricultural Land
General Disposition
Forms of Land Titling
Simple Adjudication
Title six. Agrarian Procedures
General Principles
Oral Agrarian Process
In general this legislation provides a format for the improvement of land titles in Bolivia,
20
creating the National Institute of Agrarian Reform (INRA) to oversee the process. The INRA has
a national headquarter, departmental offices, and regional judges. As can be noted in Title 2 above
there are nine components of the government structure directly involved. However, as the principal
components of the effort to improve land titles the INRA and the Superintendencia Agraria have the
following, summarized, attributes:
* Direct, coordinate, and implement policies, plans and programs of distribution, regrouping
and redistribution of lands.
* Propose, direct, coordinate, and execute the policies of community needs along with national
populations.
* Distribute land titles in the name of the Servicio Nacional de Reforma Agraria.
* Create technical direction to coordinate the old cadastral lines with local municipalities.
* Determine available agricultural land area.
* Expropriate agrarian funds (lands) for regrouping and redistribution.
* Revert lands to the state as appropriate.
* Determine and approve land area available for distribution.
* Promote and assist in the cessation of land conflicts.
* Maintain a registry of land distribution.
* Assist in infrastructure development.
* Certify land use.
Superintendencia Agraria
* Regulate and control by applying legal norms, use of natural resources.
* Dictate norms and policies for land use and classify them for highest and best use.
* Authorize the setting-aside of land for an environmental purpose.
* Assist in the expropriation of lands.
* Create and maintain a database of soil use.
* Execute inspections of soil use.
* Compile preventive measures against improper land use.
* Delegate, while maintaining responsibility, pertinent functions to the departmental or local
levels.
* Determine the amount for simple adjudication presented in this law.
* Fix the market value of land in order to provide adequate indemnity in the case of
expropriation.
* Present methods for regulation and internal control.
The Land Tenure Law of 1996 creates the National Agrarian Reform Institute (INRA) with
sole authority over the land titling process in rural areas. The Law provides a space of 10 years to
"sanitize" current rural land titles. Indigenous lands receive collective titles that cannot be bought
or sold. INRA will also take steps to establish a cadastral system with precise geographical
references and a title registration indexed by parcel of land, rather than by owner.
The Law retains the right of the State to reclaim land if its use fails to serve a useful
socioeconomic function, but this concept of what is useful is now expanded to include forestry and
conservation. Whereas before the land belonged to the person who used it in practice meaning the
22
person who cleared it for farming or livestock now the land belongs to the person that pays the
taxes on it. Under the reforms, if taxes have been paid during the previous two-years, it is assumed
that the land is serving its highest socioeconomic function.
The Land Tenure Law also established the regime for land distribution; guarantees land
rights; and created the Agrarian Superintendency and the Agrarian Judiciary. Unfortunately due to
lack of funding and the huge task assigned to INRA and the other agrarian agencies, the impact of
the Land Reform Law has as yet not been felt in most areas.
Forestry Legislation
With the passage of the first forest legislation in 1974, formal use of the forest resource has
been administered through forest contracts, similar to concessions. To be granted a contract or
concession permit, applicants had to delineate an area not previously claimed, show sufficient
funding for investment in processing facilities, generate a management plan, and begin harvest.4
Concessionaires had the right to harvest the timber as described in their management plan. They did
not own the land or indeed have rights to use it in any other productive alternative. The longest term
of a concession was for only 20 years. It has been suggested that this policy format, where there is
short duration of harvest rights and a lack of land ownership, discourages long term forest
management (Vincent 1990, Richardson 1992, Zhang and Pearse 1996).
Concession contracts were given over public and private lands alike, largely without
4 The level of necessary investment was determined by the size of the concession. In practice many of the
formal registration steps were avoided. In fact, conflicting land-use claims, unregistered harvest, and
timber piracy were significant problems.
consideration to local occupants that may have traditionally utilized the forest resource, including
indigenous groups. Besides concessionaires, hundreds of independent sawmills continued to
operate, purchasing logs at the sawmill or hiring logging crews to harvest wherever timber could be
found. Of course these independent loggers found themselves in conflict with concessionaires.
Other problems were noted in this system, including overlapping claims and oversized concessions
for land speculation purposes.
Numerous factors discouraged concessionaires from managing the resource, including the
short duration of harvest rights in concessions. Other factors were related to large economic margins
on sawn wood of a few high value species along with limited processing capabilities that would have
allowed lumber companies to add value to other species. Lack of secure tenure over concessions
certainly also discouraged operators from taking responsibility for sustainable management of the
forest. Instead, sawmills concentrated on selectively logging the highest grade timber over extensive
areas. Therefore, contrary to requirements in the law, virtually no sustainable forest management
was practiced. The role of the national forest service was limited primarily to regulation of forest
products, which unfortunately degenerated into revenue collection carried out through a large
number of road checkpoints.
In response, the government of Bolivia recently passed a new forest law5 which has
drastically altered the dynamics of the forest sector. In a bold legislative stroke, forest taxation was
changed from an ad valorem tax based on extracted timber value, to an annual, area-based tax on
5 Nueva Ley Forestal, Ley Numero 1700, del 12 de Julio de 1996. With every major piece of legislation
in Bolivia, there are associated regulations. These provide detailed instructions on how the law is to be
carried out. In this case the document is "Reglamento de la Nueva Ley Forestal, Decreto Supremo Numero
24453, del 21 de Diciembre de 1996".
hectares under concession. The area-based tax functions as a property tax and is expected to increase
economic efficiency because it has less impact on marginal production conditions than an ad-
valorem tax (Just et al 1982, von Amsberg 1994). In addition, the area-based tax system has an
overwhelming advantage in its ease of enforcement, and is easily adjusted to become more efficient
(Hyde and Sedjo 1992). Further changes in the new law include lengthening concessions terms to
40 years and allowing concessions to be traded. So far, the effect of this law has been an immediate,
drastic reduction in the area under concession in the Department of Santa Cruz from 14.7 million
hectares to 4.3 million hectares6.
It is possible that the concessions were returned to the government because they were unable
to produce sufficient revenues to cover the area-based tax. Also, the concessionaires may have
returned lands that had conflicting land use claims with, for example, colonists or cattle producers.
The result, however, is unambiguous. The government of Bolivia has suddenly regained control
over a large area of forest land and is faced with several important questions, primary of which is
as follows will the old format of selective mahogany extraction continue to provide economic
incentives for forestry under the new regime as it has for the past twenty years? In response, the
research presented here quantifies factors that have driven change in mahogany exports during the
period 1977-1993 and discusses policy options that would encourage forestry as a land use option.
There have been several changes in legislation affecting the forestry sector since the passage
of the 1974 General Forest Law (Table 1), culminating in the new forestry law and its associated
6 The newly formed Superintendencia Forestal has control of the returned concession area with part of
their mandate to reallocate as much of it as possible to forest concession.
regulations, passed in July and December 1996, respectively.
Table 1. Selected environmental and forestry-related legislation since 1974
Number Date Comments
Decreto Ley No.
116868
Decreto Ley No.
183204
Resolucion
Ministerial 380/88
Decreto Supremo
22407
Decreto Supremo
22647
Decreto Supremo
22710
Ley No. 1333
Ley No. 1493
Decreto Supremo
23660
14/08/1974 Ley General Forestal created Centro Desarollo Forestal (CDF) as an
independent administrative body under the auspices of the Ministerio de
Asuntos Campesinos y Agropecuarios (MACA).
21/02/1977 Reglamentos de la Ley General Forestal detailed the applications of the
Ley General. Includes structure and mandate of CDF; support for forest
activities; forest classification; hervest regimes and taxes; industrial
forest management and concession terms; and creates a commission on
land use.
09/12/1988 Created Subsectretaria de Recursos Naturales Renovables y Medio
Ambiente under the MACA.
11/12/1990 Established "la Pausa Ecologica Historica" for five years during which
time, no new concessions were to be granted.
Dec. 1990
Created Fondo Nacional para el Medio Ambiente (FONAMA) with the
aim of providing a funding source for environmental projects.
18/01/1991 Created the Secretaria General de Medio Ambiente under the President
of the Republic. Designed to act as an overseer of environmental issues,
replacing those previously under MACA.
27/04/1992 Ley de Medio Ambiente the objective of which is to promote
environmental protection and conservation.
17/09/1993 Ley de Ministerios y Poder Ejecutivo MACA is disbanded and the
duties are assigned to the following new ministries: the Ministerio de
hacienda y Desarollo Economico (MDHE) to promote production in the
agricultural, forestry, agroindustrial, and fishery sectors. The Ministerio
de Desarollo Sostenible y Medio Ambiente (MDSMA) to
overseeenvironmental protection, watershed protection, and multiple use
of natural resources.
12/10/1993 Regulations for the Ley de Ministerios y del Poder Ejecutivo which
designated the mandates and control of the various agencies. The
MDSMA is to establish the rational use of the natural resources. Under
the control of the MDSMA are several government bodies such as the
CDF; FONAMA; the National Institute of Colonization (INC); the
National Coucil on Agricultural Reform; and the Executive Program of
Land Rehabilitation of Tarija (PERTT). In addition, the MHDE, through
the National Secretary of Agriculture and Ranching and the Subsecretary
of Forestry and Fishery Development, will work in coordination with the
MDSMA to promote agricultural and related production.
Ley No. 1700 12/07/1996 Ley Forestal replaces 1974 forestry law. Abolishes CDF and replaces it
with Superintendencia Forestal to oversee all aspects of the forestry
sector. New tax based on concession area rather than volume extracted.
Decreto Supremo 21/12/1996 Regulations of the new forest law. Among other objectives it determines
24453 mandate of Superintendencia Forestal and funding sources.
Source: Lopez (1994).
The stated objectives of the law are to normalize the protection and sustainable use of the
forest for the benefit of current and future generations, while reconciling social, economic, and
ecological requirements. Several principal components of the legislation are readily apparent, such
as the creation of the Superintendencia Forestal (SF) as the autonomous body in charge of overseeing
the forest sector. The SF has the following, summarized, mandates:
* Supervise the forestry sector of the nation.
* Authorize and supervise concessions and other forest use plans.
* Register concessions and ecological reserves.
* Control illegal forest activity.
* Complete necessary inspections and apply appropriate penalties to ensure compliance with
current law.
* Conduct external forest audits, make results and remedies available.
* Collect, verify, and distribute revenue from forest taxes.
* Delegate functions to municipalities according to ability of a said municipality.
* Control resources under the SF administrative process.
This newly created agency, under the auspices of the Ministry of Environment and
Sustainable Development (MDSMA), supplants all previous government forest agencies, and has
a similarly broad mandate as the Superintendencia Agraria. The SF is solely responsible for the
management of the concessions, and for the control and distribution of an additional eight million
hectares of land (in Santa Cruz alone) which has been returned to the state. Funding for the
Superintendencia Forestal comes from one source, the patente forestal (or area-based forest tax).
Set at $US 1.00, this tax will provide approximately five million dollars of revenue annually.
However, the SF only receives 30 percent of that total, or 1.8 million dollars, as an operating budget
for management of the entire forest sector of Bolivia. Thus, as is often the case with directed taxes
of this nature, the SF is encouraged to distribute as much land to forest use as possible to improve
its revenue. However, land distribution will not be easy because (a) one could argue that the land
that has been returned is unproductive or rather has economic constraints to extraction, and (b) there
is some discussion that much of this land actually is on private lands or has another conflicting land
claim which will hinder redistribution.
Another important component of this legislation is the change in forest taxation methods.
Previously a volume-based tax on timber extraction, the taxation system is now an area-based tax
extracted on a per hectare basis. Theory suggests that in even-aged forest management, area-based
taxes reduce the rotation length in forestry production, whereas volume or yield-based taxes lengthen
the rotation. The equivalent in tropical forestry would be an increased intensity of management or
extraction per hectare with an area-based tax to cover the fixed annual outlay. Although not
theoretically as efficient at rent capture as sealed bid concession allotment, or even the previous
volume-based tax, the area-based tax system has an overwhelming advantage in its ease of
enforcement (Hyde and Sedjo 1992), which was the overwhelming argument for its adoption by
28
Bolivian authorities.
With the dramatic decrease in the area under concession from 21 million hectares before the
new law to about six million nationwide today, the impact of the area-based tax has become readily
apparent. One interpretation of the massive reduction of concessions is that the present value
expected from timber extraction is insufficient when compared to the annual area-based tax, which
is set at a minimum of $1.00/hectare/year over the entire concession area. In fact a much lower tax
of $0.40 was recommended by Hyde et al (1996) who based their recommendation on 10 percent of
an estimated $US 4.00 per hectare annual stumpage value.
Private forest lands pay the same patent per hectare as concessions, but only over the area
of annual allowable cut. In other words, private forests under management plans pay only one dollar
per hectare on the area harvested, rather than the entire productive forest area. With a minimum
cutting cycle of 20 years, forest landowners need only pay $1.00 for every 20th hectare, or an
average of $0.05 per hectare of forest per year. With longer cutting cycles, the tax per hectare of
forest is even lower.
An interesting application in the new law has been the lifting of the ban on exporting logs.
There is a consensus in international development literature that bans on log exports can, rather than
preventing deforestation and increasing value-added, cost the government money. With this in mind,
the new forest legislation has repealed the log export ban. Caveats to log exports, however, are that
the logs must be of a lesser known species, the log itself must be more valuable than the processed
wood, and lastly the export ban has been lifted for only five years. Combined, these factors may
constrain the potential benefits of this legislation.
Under the new Forestry Law, the benefits of harvesting forest products now accrue to the
29
landowner or concessionaire. In order to be harvested, forests must be managed under an approved
sustainable forest management plan, which requires a forest inventory, definition of cutting cycles,
details on regeneration, experimental plots to be measured at five-year intervals, logging practices
to minimize damage, measures to protect waterways and wildlife, and such.
All harvest of forest products requires an approved sustainable forest management plan,
whether on a concession, a tract of indigenous lands, or a private property. The only exception to
this requirement is harvest during land clearing, which is approved under an individual property land
use plan.
Forest concessions are now awarded only on public lands. Private land ownership has
priority over public concessions, and new concessions are given only through international bid.
Other important changes in the forest law were that concession duration was lengthened to 40 years,
but renewable every five years, and concession rights can be traded. It is hoped that these changes
to more secure tenure will induce greater investment in silviculture and improved forest management
(as suggested by Zhang and Pearse 1996).
Another initiative in the forest law is an effort to decentralize the control of forest use and
management by incorporating local municipalities. Individual municipalities are to be awarded 20
percent of the public forest land within their jurisdiction to be maintained in forest production. In
a recent study of the changing role of municipal governments and forest management in lowland
Bolivia, Kaimowitz et al (1997) conclude, but not without reservations, that strengthening local
government roles in forest management will be beneficial in natural resource management.
Further reforms in land taxation
Under the new regime, property taxes are established for grazing and forest land, plus a
specific tax for land clearing. Ranch land incurs a progressive tax according to the total land value
declared by the owner on a progressive scale beginning at 0.35% and increasing to 1.50% of the
value without improvements (Table 2).
Table 2 Land Tax Schedule for Livestock Ranches, 1996
Declared value Fixed tax Plus variable On an amount
per hectare (USD) rate (%) over (USD)
$0 $40,856 0 0.35 0
$40,856 $81,712 143 0.50 40,856
$81,712- $122,569 347 1.00 81,712
>$122,569 756 1.50 122,569
Actual land prices for cleared pasture land may be as high as $500 per hectare in accessible
areas along major roads in regions such as the Chiquitania of Santa Cruz. Prices decrease
dramatically outside these zones, and drop to $20 or even lower for uncleared land without title
toward the Brazilian border. Generally, pasture lands with limited access may be valued between
$100 and $250 per hectare. Depending on the size of ranches and the total value, taxes on land
cleared and planted to pasture are typically between $0.20 and $0.50 per hectare.
Under the new law, land clearing is taxed on both area and volume of timber removed. The
area tax or patent is currently set at 15 times the per hectare patent paid be forest concessions, or
$15/hectare.7 The volume tax is computed at 30 percent of the value of timber at the land clearing
site, according to a price list established by the Forest Superintendency. The landowner pays 15
percent of the value and the buyer of the timber another 15 percent. The price list applied by the
8 The forest concession patent was initially set at $1 per hectare. It can be adjusted upward in future years
with changes in international tropical timber prices, but can never be lower than $ 1/hectare.
31
Forest Superintendency for 1998 seems to reflect international timber prices less extraction and
milling costs, which are several times higher than the amounts paid by local sawmills. Therefore,
timber will be extracted from land clearing sites only in cases where the landowner is able to find
interested lumber industries with capacity and access to international markets. Otherwise, the timber
will be burnt off, a practice common in the Bolivian lowlands.
Land use planning requirements
The current Forestry Law requires preparation of "Land Use Plans" for individual properties
to be prepared within the context of regional land use plans (locally referred to as PLUS) currently
being developed for the lowlands of Bolivia. Land clearing is to be approved only according to
individual land use plans.
Individual land use plans include the concept of ecological reserves in areas for windbreaks,
close to streams, on steep slopes, in wetlands, on thin or rocky soils, etc. Land in these areas cannot
be cleared of natural forests, and no forest harvesting activities can be carried out in these reserves.
Individual land use plans also specify the uses that can be made of the land resource, classifying
areas for intensive annual cropping, perennial crops, pastures, open grazing, and forestry. The plans
prescribe minimum conservation practices to be required ofthe owner/operator. Individual property
land use plans are to be registered with land titles and become legal zoning restrictions or easements
applying to current and future owners.
Just as with sustainable forest management plans, individual property land use plans are
prepared by professionals hired by landowners, but who are legally responsible for the veracity of
the information and analyses to the Agrarian and Forest Superintendencies.
32
The cattle and forestry conflict in Latin America and implications for Bolivia
For the past two decades there has been extensive debate on the destructive interaction
between cattle production and tropical forests in Latin America. The primary concern has been
deforestation to provide land for pasture and cattle production. The literature has focused, with
good reason, on two geographic regions: Central America, where the concern was that forest were
being destroyed to produce cheap beef for the US markets; and Brazil, where vast areas of the
Amazon were opened up for cattle production. Although these two areas have received much of the
attention, the interaction of cattle production and forests is apparent in many countries of Latin
America, including Bolivia.
The increase of land allocated to cattle production at the expense of natural forests is the
result of several interactive factors9 which include: land tenure policies (Jones 1990); extensive
cattle production systems (Serrao and Toledo 1992, 1993); low timber values (Kishor and
Constantino 1993); production characteristics of cattle (Hecht 1992); favourable markets and prices
(Meyers 1981, Nations and Komer 1983); and government subsidies for livestock credit and road
construction (Mahar 1988, Binswanger 1991). Of these characteristics, the first four are applicable
to Bolivia. Historically, land tenure has been established by clearing. As with many other countries,
this has lead to uneconomic exploitation and destruction of the forest resource. This situation has
changed however with the inception of the new agricultural law which described forestryas an "Uso
Mayor", or productive use. This simply states that forestry is a productive use, sufficient to establish
8 The term "hamburger connection" was coined by Meyers (1981) to describe this phenomenon.
9 The list is taken from Godoy and Brokaw (1994) cited in Kaimowitz (1996).
33
tenure. Therefore, in theory, Bolivia has made a substantial change. In practice, the impact has not
yet been felt.
The second factor, slow technological change, has been a truism of cattle production in
Bolivia. Production is mostly extensive and Bolivia still has not controlled Foot and Mouth disease.
There is however, an increasing herd, located in the area immediately surrounding the city of Santa
Cruz, that is under intensive management, and these techniques may transfer to other producers in
the future. Third, Bolivia has not actively had policies that depress timber values, but legislation that
differentiated between sub-soil resources, the soil resources and above-soil resources, has served to
make timber extraction difficult for private landowners. The legislation stated that sub-soil resources
(minerals) and above-soil resources (forests) belonged to the state and were allocated by extraction
permits; soil resources, on the other hand were allowed to be held in private ownership. The
confusion arose regarding the extraction of timber from private land. To remedy this predicament
the current legislation allows extraction of timber from private lands. Unfortunately, the land use
plans required for such an activity are prohibitively expensive for the average landowner (see chapter
3). Fourth, in Bolivia, the characteristics of cattle such as: low labor and supervision inputs,
transportability, limited risk, prestige, limited inputs, and biological and economic flexibility, all
play a role in the decision to participate in production. In addition to the three factors above serve
to define many of the reasons why the cattle sector of Bolivia has maintained a steady presence for
many years.
The final two explanations are not applicable to the growth of cattle production in Bolivia.
That is to say, these factors have not been part of the economic environment and therefore have not
played a significant role. The market for Bolivian beef, rather than international, is limited to
34
domestic consumption. Finally, there has been no subsidized credit to cattle production nor have
there been substantial improvements in infrastructure. However, land use in the Bolivian lowlands,
and in particular the cattle and forestry industries, is in the process of radical change.
Cattle production in the Bolivian Lowlands
Cattle ranches in Eastern Bolivia are usually extensive operations, characterized by low input
and management costs. Stocking rates and production efficiency have been notoriously low, and
there have been few technological changes until very recently. Although beset by problems, land
in the Bolivian lowlands is suitable for cattle production--a fact confirmed by the land use plan for
the extensive Santa Cruz Department, which shows 23.6 million hectares (64 percent), of a
Department total of 37 million to be.suitable for cattle ranching in one form or another. The cattle
sub-sector contributed 7.5 percent to the 1998 regional value product, with a total of 56.5 million
dollars.
The current Bolivian cattle population is approximately six million head, up 9 percent from
the 1990 population of 5.5 million. Of the six million head, 2.7 million are in the Beni and 1.5
million in Santa Cruz, such that these two departments alone comprise 70 percent of the total herd.
In the center of the Department of the Beni, livestock range on vast natural grasslands. This region
is subject to severe seasonal flooding in summer months (November through March) and to dry
conditions prevailing in winter. Ranching in this region is extensive, but the natural grasslands still
permit cattle production at low cost. Ironically, the major production constraint in this region is not
the ability to find feed in the dry season, but rather to find high, dry land in the wet season. In the
past, cattle from the Beni plains were driven to the abattoir in Trinidad, and the carcasses were then
35
transported by air to La Paz.
The Department of Santa Cruz covers more than one third of the country's entire territory,
and lies at the center of the lowlands, ranging from humid areas to the north bordering the Beni to
the dry Chaco in the south. The three principal regions of cattle production within Santa Cruz depict
the variations found throughout the east. The first is the integrated or expansion zone surrounding
the city of Santa Cruz where intensive beef cattle and dairy production are developing alongside a
thriving agricultural industry based on sugar cane, rice, soybean, sunflower, sorghum, wheat and
other field crops. About a third the Department's cattle population can be found here under relatively
intensive management on improved grasses. In many cases, land that was originally cleared for
agricultural production has reverted to pasture. This land is accessible to the Santa Cruz market and
is virtually all cleared for cultivation, however, it is relatively high price a combination of attributes
which encourages more intensive management. In recent years, it has become common to bring
young steers into cultivated pastures near Santa Cruz for fattening before final sale, frequently at
stocking rates up to two animals per hectare. It is expected that the trend toward cattle production,
especially growing/fattening operations, will continue to expand in this area. Marginal agricultural
land is available for productive establishment of pasture. In addition, the burgeoning urban market
of Santa Cruz, now close to one million, will to provide a substantial demand.
The second production region is the Chiquitania which cuts a diagonal swath from top left
to bottom right through eastern Santa Cruz. Designated as mostly joint, sustainable forest
management and limited cattle production, this area holds about 40 percent of the cattle population
of Santa Cruz (Table 4). Extensive ranching on wooded savannas characterizes cattle production
in much of the area, and the interaction between forests and cattle has been underway for more than
36
two hundred years (Kileen 1991). Cattle production is increasing in this region, with the gradual
clearing of forests by hand and sometimes using mechanized methods. Most soils are not apt for
agricultural production, while infrastructure improvements have made the Santa Cruz market more
accessible for cattle ranchers. Improved grasses are a necessity for cattle production here as
domestic grasses are of poor quality.
The final area of cattle production is the Chaco where production is practiced on an extensive
scale, with as much as 15-20 hectares needed to maintain a single animal unit. This harsh, dry region
holds approximately 20 percent of the Departmental herd, and much of this region is described in
the Departmental land use plan as useful only for extensive cattle production, with some forestry.
The Bolivian Chaco region continues further south through the Departments of Chuquisaca and
Tarija to the borders with Paraguay and Argentina.
Table 2. Cattle Population in Santa Cruz by Province, 1992-1995
Province 1992 1993 1994 1995
(head)
Andr6s Iban6z 126,916 130,723 134,645 138,685
Warnes 68,403 70,455 72,569 74,746
Obispo Santiestevan 43,516 44,821 46,166 47,551
Sara 86,222 88,809 91,473 94,217
Ichillo 48,722 50,184 51,689 53,240
Florida 76,123 78,407 80,759 83,182
Vallegrande 108,386 111,638 114,987 118,436
M.M. Caballero 40,584 41,802 43,056 44,347
Cordillera 239,134 246,308 253,697 261,308
Chiquitos* 110,360 113,671 117,081 120,593
Rluflo de Chavez* 117,686 121,217 124,853 128,599
Velasco* 151,568 156,115 160,798 165,622
Angel Sandoval* 181,532 186,978 192,587 198,365
German Busch* 53,158 54,753 56,395 58,087
Guarayos 39,335 40,515 41,731 42,982
Total 1,491,645 1,536,396 1,582,486 1,629,960
37
* Provinces which comprise the Brazilian Shield/area of study (41% of the total herd, 1995).
Source: CAO 1996
This chapter will illuminate some of the constraints and potential in the Bolivian cattle sector
by addressing the following: markets and prices; production activities, a timetable and budget; and
finally, grasses, land clearing and fencing.
Beef markets and prices
Limitations in the export markets are created by high transportation costs, and the continued
incidence of hoof-and-mouth disease. Until a few years ago, much of the meat supply reaching the
city of La Paz was transported by air from rustic slaughtering operations at ranches in the Beni.
Cattle are still shipped by river in the Beni and then trucked to market in Cochabamba. Even
transportation to the Santa Cruz market is difficult from many areas. The only three commercial
abattoirs are located within the city of Santa Cruz, and therefore animals are transported live from
connecting regions. Efforts to export to points east and south (Brazil and Argentina) are stymied by
lower prices and higher product quality in those regions.
The second export constraint is hoof-and-mouth disease, which creates an export barrier to
countries or areas free of this disease. There is an active campaign to eliminate hoof-and-mouth
disease from certain areas in the Bolivian lowlands, but meanwhile the severe constraints on exports
ensure that the domestic market for beef is of paramount importance to producers.
Local prices are cited in dollars per kilogram on the hook. Animals are usually grown out
on ranches and are marketed at full size between three and four years of age at about 350 to 400
kilograms live weight. Although national consumption is very low, only about 18 kilograms per
capital annually, average beef prices have risen from to about $1.50 per kilogram at present with little
price differential is given for higher quality beef.
Production Activities, Timetable and Budgets
Production activity for cattle is aligned with the seasons. Calves are timed to be born in
August, immediately before the rains, and to be weaned in April/May at 7-8 months old. All
associated activity is based around this cycle. Table 4.2 shows the production timetable for the
experimental ranch "El Salvador" located in the Chaco (Saravia et al. 1995).
Table 3 Schedule of cattle production activities
Affliction and control Month J F M' A M' J J A S N D
Foot and Mouth--once/yr, all herd
Rabies--once/yr, all herd
Blood infections--all herd
I I I I I I I I I
Gangrene--all herd
Neonatal diarrhea--birthing months
I I I I
Brucelosis--all females 1-2 years old I I
Internal (ecto) parasites injection =
i I I I A I I I I
External (endo) parasites -- all year A A A
Trichonomas, Leptospirosis, and A
I I I I A A A A I
Baloneprostitis -- in bulls
Mineral salts--during rainy season I A I
I I I A
Neumoenteritis--alcohol at birth A I I I
Source: FEGASACRUZ: Manual de Ganaderia del Chaco.
Within this timetable, strict attention should be paid to pest and disease management under
the difficult conditions of the Bolivian lowlands. However, these practices are often neglected, to
the detriment of production efficiency. The following is a brief description of the pest and diseases
prevalent in the Bolivian Lowlands, and indeed throughout much of Latin America.
At birth, calves should receive injections against endo/ecto parasites. Worming is then done
about every three months thereafter until calves attain about 300 kilograms. Cows and bulls should
be wormed every six months at the beginning and end of the rainy season. External parasites such
as ticks, mosca del cuerno, and borboro Dermatobia hormini are controlled by spraying with
insecticide. Regular rotation of pastures also helps in tick control. Calves should get three vaccines
at three months old--against Black leg (gangrene), Malignant edoema, and Hemmoragic septiciema.
Female calves receive and additional vaccine at 3-7 months against Brucelosis. All animals are
vaccinated against rabies once a year and hoof-and-mouth disease once a year. All females are
vaccinated once a year against Leptospirosis.
There are economies of scale apparent in cattle production in Bolivia. Total production cost
on small, medium, and large ranches are estimated to be around: $1.13/kg, $0.72/kg, and $0.61/kg,
respectively (Table 4.3). Fixed costs of production, which include administrative costs and
depreciation costs, range from: 60 percent for small, 56 percent for medium, to 50 percent for large
producers.
Table 4 Cattle Production Budgets for the Chiquitania (San Ignacio San Jose de
Chiquitos)
Description Small Medium Large
Hectares 500 ha. 1,000 ha. 2,500 ha.
Number of cows 250 cows 610 cows 1,200 cows
Beef production 13,812.5 kg 33,702.5 kg. 66,300.0 kg.
Cost description $ total $/kg. $ total $/kg. $ total $/kg.
Fixed costs 9,280 0.67 13,751 0.41 20,492 0.31
Depreciation 2,927 0.21 4,568 0.14 7,776 0.12
Administrative 6,353 0.46 9,183 0.27 12,716 0.19
Variable costs 6,292 0.46 10,660 0.32 19,951 0.30
Labor 2,161 0.16 2,165 0.06 3,829 0.06
Replacement animals 1,080 0.08 1,880 0.06 3,480 0.05
Health 575 0.04 1,331 0.04 2,570 0.04
Miscellaneous 846 0.06 1,974 0.06 3792 0.06
Sales/marketing 1,230 0.09 2,910 0.09 5,880 0.09
Technical assistance 400 0.03 400 0.01 400 0.01
Total production costs 15,572 1.13 24,412 0.72 40,444 0.61
Notes: Totals may not add due to rounding.
Source: CAO, 1996.
Land clearing for pasture
There are several methods of land clearing, and the appropriate one will vary according to
forest type and operator preference. For lower, sparse forest typical of much of the Chiquitania the
choice is often "cadeneo" or chain clearing, which consists of dragging a large chain between two
bulldozers. At around $200 per hectare this is a very cost effective method for low forests. One
month after the "cadeneo" the land is burnt, then seeded. Seeding is done by hand, on horseback,
on a tractor, or by airplane. Four months later, the grass will be dropping seed, and cattle must be
allowed on. The land is burnt again in the second and third years to, and the field should be
effectively grassed. The schedule is roughly as follows: clear in August, burn and seed in
September/October, and put cattle on in February. However, because of the remaining stumps and
half-burnt logs, mechanical control of brush and weeds is not feasible for many years.
With taller forests, "cadeneo" is less feasible. Instead, bulldozers push trees and brush into
windows. "Dead windows" consist of piles of brush to be burnt and the area eventually grassed.
Live windows are left as standing forest for windbreaks and refuge for cattle during periods of cold,
windy weather. By pushing 30 meters in each direction, an opening of 60 meters is created. Three
such swathes, and two intervening "dead windows" of 15 meters each give a total width of 210
meters. Then a live window of 60 meters should be left. This method, typical of that used with
mechanized farming operations, is much more expensive and "cadeneo," but has the advantage of
allowing mechanical control of weeds. Clearing costs can range between $300 and $550 per hectare,
depending on the density of the forest, the remoteness of the region, and the size of the area to be
cleared.
Costs of clearing using mechanized "acordonado" with a bulldozer and blade pushing brush
and trees into windows are estimated at about $388 per hectare for tall forest with about 65 m3 of
timber, plus about $13 per hectare for transport of equipment. Clearing also involves burning of
"dead windows" and cleaning of borders for fencing. Costs vary according to the density of the
forest.
Other methods are possible, where trees are cut by hand, possibly using chainsaws. Hand
clearing implies leaving stumps, which for tropical hardwoods will remain in the ground for more
than 20 years, complicating weed and brush control. The first option is most often used for cattle
42
areas. There is a problem, however, in that it may be more costly to clear smaller areas. I used 200
hectares in these examples, but even the base price of $120 may not hold for areas less than 500
hectares of more. Transport would be higher for smaller areas.
Landowners can benefit from harvesting timber from the areas being cleared. Forest
inventories in the management plans provide an estimate of the commercially available timber--an
amount greater than the Annual Allowable Cut per hectare in that the entire volume of commercial
timber is included, with no margin for minimum diameter or seed trees. Revenues are estimated by
taking these available volumes times current stumpage prices shown above, less volume tax.
Estimations of revenues from sale of timber at time of clearing range from only $7 per hectare in the
Chaco to as much as $189 per hectare in Guarayos. The net cost of clearing and establishment of
fenced pastures in the base analysis ranges from $242/hectare in the Chaco to $489 in Bajo Paragud.
Clearing also requires the presentation and approval of an individual property land use plan
and land clearing plan. Associated harvest of timber requires issuance of "certificates of forest
origin" with payment of the area and volume-based taxes as described in the following section.
Under the new law, land clearing is taxed on both area and volume of timber removed. The
area tax or patent is currently set at 15 times the per hectare patent paid be forest concessions, or
$15/hectare.10 The volume tax is computed at 30 percent of the value of timber at the land clearing
site, according to a price list established by the Forest Superintendency. The landowner pays 15
percent of the value and the buyer of the timber another 15 percent. The price list applied by the
Forest Superintendency for 1998 seems to reflect international timber prices less extraction and
The forest concession patent was initially set at $1 per hectare. It can be adjusted upward in future years
with changes in international tropical timber prices, but can never be lower than $ 1/hectare.
43
milling costs, which are several times higher than the amounts paid by local sawmills. Therefore,
timber will be extracted from land clearing sites only in cases where the landowner is able to find
interested lumber industries with capacity and access to international markets. Otherwise, the timber
will be burnt off, a practice common in the Bolivian lowlands.
Fencing is estimated to run about $376 per kilometer, assuming the terrain has already been
cleaned when the land was cleared, and that posts are available on the property. Wire in 500 meter
rolls requires eight rolls/kilometer when using four strands stapled to posts. With pastures of 50
hectares, the average cost of fencing per hectare comes to about $21. Establishment of pasture
requires approximately another $40 for seed and $20 for seeding. Other investment costs for items
such as water systems, corrales, and infrastructure for housing are not considered in the study. An
average annual fixed cost of $ 1,000 is included to reflect these costs, besides the annual maintenance
cost per cow of $12.62.
In areas closer to markets, timber might be sold before burning, but as stated in the previous
chapters, local markets for common species used in construction are limited.
Of the wood products used on-farm, by far the most important is the use of trees as fence posts. The
logs are quartered, and depending on the size of the trunk will give anywhere from four to 10 posts
per tree. Large posts machiness) are placed every 100 meters at 75 cm deep, and smaller ones every
5 meters at 60 cm deep. In this manner, 200 posts are needed per kilometer. The most popular
species are very hard woods such as Cuchi (Astronium urundurueva).
Grasses
With the exception of the Beni plains, improved grasses are an important component of cattle
44
ranching in much of the lowlands. In Santa Cruz, a total of 61,090 hectares has been planted to date
(CAO 1997). Although there have been substantial reports of declining pasture productivity in
pasture created from tropical forests in Brazil (Buschbacher et al. 1988), in most of the areas of
Bolivia, it is accepted that grasslands can remain productive over long periods of time and, in fact,
suffer more from problems of weed infestation rather than declines in soil productivity (Kileen
1991).
The most popular grasses are of the genus "Brachiaria." Brachiaria deCumbens performs
best on well-drained soils and is considered by some to be the best all-around grass for much of the
Bolivian lowlands--because of its ability to resist trampling and provide forage during the dry
season. "Bracharion" (Brachiaria Brisantha) is also used, but does substantially worse where
drainage is limited. Brachiaria humidicola will tolerate both well-drained and poorly-drained soils
and can survive in areas of standing water for long periods. It is also tolerant of dry conditions as
are the other two Brachiarias. However, it does not have as high a nutritional value as the other two
and is therefore good for breeding and growing, but not for fattening. Tanner grass (Brachiaria
radicans) performs well in areas subject to flooding, and will tolerate dry conditions, but must be
planted by stolon, rather than seed, which is labor intensive. Other Brachiara species which are
becoming commercially available are Brachiaria dictoniana and Brachiaria ruziziensis.
Another popular grass is YaraguB (Hyparrhenia rufa) which especially predominant in the
Chiquitania. It is palatable when young, and very vigorous on well drained land, such that it will
tolerate poor land if drainage is adequate. It does not tolerate dry conditions and if not properly
managed can become lignified. Finally, Guinea grass (Panicum maximum) and assorted
improvements (Tanzania, Tobiati, Centenario) are thought to be among the best fattening grasses
45
in South America. They need well-drained soil but do not do well in dry conditions. These grasses
can often be seen in cut-and-carry operations for breeding stock.
Table 7 Characteristics of Selected Grasses
Min Cold Drought Flood Soil Percent Prod.
Grass Precip. tolerance resistance tolerance quality protein (MT/ha.)
B. brisantha 1,000 low medium low low/med 6.29 50
B. decumbens 800 low medium low/med med 4.1 60
B. humidicola 1,000 medium med/high med/high low/med 2.3 45
B. ruziziense 1,000 low low low med/high 6.29 50
Coloni6n 1,000 low low/med low high 12.2 50
Tanzania 800 low low/med low high 12 60
Tobiati 800 low low/med low high 10 60
Source: Sementes California
Cattle and beef transportation
Most of the beef in Bolivia is transported live, either by herding, truck, rail or river.
Shrinkage, or weight loss during transport, is an important factor with these kinds of transport. Only
recently, ranchers in the La Paz-Beni region have begun slaughtering locally and shipping beef sides
to La Paz in refrigerated trucks.
Costs from most regions are quoted from a central shipping point, so there is an additional
charge to bring cattle from the ranch, either by herding or trucking. Only in Guarayos were quotes
obtained for shipment directly from the ranch to market.
Cattle are not currently shipped from the Pando. The region has so few livestock that it is
actually a net importer, from the northern areas of the Beni. For purposes of testing the model,
however, the costs of shipping beef from the region to the most accessible external market were
included. Whereas lumber from the region would be shipped upriver Puerto Villaroel and by truck
to Cochabamba, this route is impractical for shipping live cattle because of the time required."
Alternatively, cattle could be trucked during the dry season to the San Borja area in the La Paz-Beni
region, slaughtered there, and the beef shipped by refrigerated truck to La Paz. These figures are the
ones applied in the model.
Shrinkage is highest for the Pando and Bajo Paragud regions which involve shipping live
cattle by truck for distances of 700 kilometers during three to four days. Next are the Chiquitania
and Chaco which involve two days of shipping by rail or truck. The lowest is the Guarayos region
from which live cattle are shipped easily in one day by truck, and the La Paz-Beni region where beef
is slaughtered locally for shipment to La Paz.
Finally, the cattlemen's associations and municipal governments charge taxes on animals
entering slaughterhouses. The figures' uses are typical charges, provided by the Federation of
Cattlemen in Beni and Pando.
11
Further upriver beef is commonly shipped from the Beni plains into Puerto Villaroel and on to market in
Cochabamba.
Table 8. Transportation costs for cattle and beef.
Gran Bajo La Paz
Chaco Chiquitania Guarayos Paragua Beni Pando
(USD per head)
Herd/truck 8.00 8.00 8.00 8.00 8.00
Rail 16.22
Truck 36.36 25.00 52.50 52.50
Referated truck 34.48 34.48
Truck to Market 5.00
Federations 2.91 2.91 2.91 2.91 2.91 2.91
Municipal Tax 7.27 7.27 7.27 7.27 7.27 7.27
TOTAL 54.55 39.40 35.18 70.68 55.66 105.16
The Forestry sector: description and issues
There are six identified forest ecozones throughout Bolivia (Lopez 1993). These are:
Evergreen lowland forest found principally in the Beni and Pando, to a lesser extent in Santa Cruz,
Cochabamba, and La Paz. This formation is extensive (approximately 245 Km2) throughout Northern
Bolivia, and is characterized by undulating alluvial plains lower than 500 meters above sea level
(masl), annual rainfall 1,300 to 2,300 mm and with a 2-3 month dry season. Moving south, the Sub-
tropical humid forest in Santa Cruz is the most important commercial zone, recording an average
of 60 percent of official total log production and 80 percent of mahogany exports during the period
1980-1993; accounting for a total export value of $ 44.5 million in 1994 (CNF 1995). This zone is
characterized by a diverse forest with over 100 potential wood species, on a flat plain, at 300-400
masl, with an annual rainfall of 1,200 to 2,000mm. The Semi-humid low forest is found in the east,
mostly in the Chiquitania region of Nuflo Chavez and Velaso provinces. Annual rainfall for this
zone is 900 to 1,200mm and the topography is fairly flat between 350-400 masl. The Semi-humid
mountainous forest extends south, is less productive, and reaches as high as 2000 masl. The Semi-
arid low forest is found in the Southwest of the forested region and has lower forests. Rainfall varies
between 500 to 1000mm and altitude ranges from 350 to 1000 masl. The Yungasforest is found in
the northwest of the forest region, in the valleys of Cohabamba and la Paz (Lopez 1993).
There are conflicting claims of forest productivity however, in as much as there is a range
of estimates from 12 to 18 million hectares of productive forest (Mancilla 1996, ITTO 1996,
CORDECRUZ). Santa Cruz has been the most important area of log extraction, with approximately
54 percent of total log extraction in 1995, however, as commercial stocks become depleted, more
attention has been given to other areas. This was demonstrated by the recent granting of 19 new
concessions in the Pando, a region where there had previously been little forest activity.
Table 9. Logging by Department, 1995
Department Volume of logs (m3) Percent
Santa Cruz 240.54 53.61
Cochabamba 91.88 20.48
La Paz 73.21 16.32
Beni 39.07 8.71
Other Departments 4.00 0.89
Total 448.70 100.00
Source: CNF(1995)
Export markets
The use of the forest resource in Bolivia, has, in the past ten years been expanding rapidly,
principally to provide for the export market which is estimated to be 70 percent of the total market.
Wood exports were valued at 82 million USD in 1996, up from 49 million in 1990. The driving
force for the past twenty years in the forestry sector has been the selective logging and export of
49
mahogany. There is a genuine concern over the stock of mahogany in Bolivia, with estimates
suggesting that there are as few as five years production remaining. Although unsubstantiated, it is
accepted that mahogany stocks are in serious decline, and with continued extraction will become
commercially extinct.
Table 10. Markets for Bolivian forest products, 1995
Final market by type of product Value in US$ Percent
Export of sawnwood 63,188,033 56
Export of semi-processed and processed wood products 15,617,689 14
Domestic markets 33,770,879 30
TOTAL 112,576,601 100
Note: Totals may not add due to rounding. Source: Pattie and Aguilar based on data reported by CAmara Forestal de
Bolivia (CFB) and the Private University of Santa Cruz (UPSA)
Table 11. Value of Forest Product Exports from Bolivia
PRODUCT GROUP 1990 1991 1992 1993 1994 1995 1996
(millions of dollars)
Sawnwood 36.4 41.6 41.2 44.5 70.8 63.2 64.6
Processed Products 12.5 7.4 8.8 8.8 18.6 15.6 17.8
TOTAL 48.9 49 50 53.3 89.4 78.8 82.4
Note: Processed products include sleepers, veneer, plywood, flooring, moldings, and others.
Source: Pattie and Aguilar, based on data from the Camara de Industra y Comercio and the CAmara Nacional Forestal
With the decline of mahogany stocks, there is now a concerted effort to improve the per
hectare value of forest land by encouraging the extraction and export of lesser known species. These
species are universally less valuable than mahogany, therefore, lowering extraction costs whether
by improved management and/or extraction costs, will be a vital aspect of the potential profitability.
To date, relatively few have been exported in any significant volume, with Roble (Amburana
caerensis) and Cedro (Cedrela sp.) as two notable exceptions. Indeed, roble has overtaken
mahogany in sawnwood export volume for 1994 with 35,000 m3 compared to mahogany at 30,000
m3, however, the mahogany value remained greater at $US14.309 million vs. $US10.916 for roble.
Currently Ochoo (Hura Crepitans) is finding wider appeal and is exported in addition to being a
mainstay in the domestic market. The importance of these alternative products becomes apparent
when examining the changes in export volumes and diversity. The most prolific of the sawnwood
exports for 1994 include the following species: Yesquero (Carinaiara estrellensis), Morado12
(Peltogyne sp.), Tarara (Centrolobium sp.), Sirari (Ormosia sp.), Tajibo (Tabebuia sp.), Plumero
(Vochysia lanceolata), Cambara (Erisma uncinatum), Palo Maria (Calophyllum braziliense), Paquio
(Hymenea stilbocarpa). The diversity of species now exported in the form of sawnwood has
increased dramatically from only eight species in 1985 to twenty-four species in 1994. Also value-
added is found in the following product categories: for semi-processed wood terciada, laminates,
parquet, machiembra, tapajunta, aglomerado, and revestimiento; processed products (in approximate
order of 1994 value exported) sleepers, doors, molding, windows, furniture, beds, guitar parts,
chairs, tables, stairs, gates, posts, and so forth. Of the sawnwood exports, principal destinations for
12 Morado sawnwood export volume is low, but at a price of 2.27 $US/bdft, which is more that
twice that of mahogany. However, morado is exported primarily as residuals (190,907 bdft @
$US2.36/bdft) which are used in handicrafts.
51
forest products during 1994 were Argentina, which received 58.24 percent of the total export value,
the US with 17.26 percent, Uruguay with 8.2 percent, Mexico with 5.87 percent, Chile with 2.16
percent, and Japan with 1.87 percent.
The Bolivian forest sector, however, is in the midst of change. Nineteen new concessions
have been allocated in the far northern Department of Pando. And, with mahogany stocks in the
decline (Stolz and Quevedo 1993; Merry and Carter 1999), there has been a concerted effort to
change from selective harvest techniques to sustainable forestry management that relies on lesser-
known (and lesser-valued) species. Indeed, Cedro (Cedrelafisilis) has recently become the principal
export wood in terms of both volume and value.
Domestic markets
The domestic wood market is characterized by the extensive use of low grade timber however
data on the market is unreliable at best. A recent ITTO estimate puts the volume at around
324,000m3. This market is primarily supplied by wood that comes from small farmers and colonists
clearing land for agriculture, or wood that is not of exportable quality from the large sawmills.
The process of domestic supply from small farms is as follows. Small sawmill owners with
wholesale shops (Barracas) in Santa Cruz act as middlemen. They look for new clearing, and will
extract the timber immediately before the process begins. This provides a service for the farmer in
that he does not have to clear away the biggest and most cumbersome trees. The stumpage value
for these trees, regardless of size and quality, is very low, often as low as $ 5-10 per trunk. The low
stumpage value is a function of a saturated market, an abundance of the resource, and a need for the
farmer to remove the tree in order to farm. The timber merchants also create rudimentary roads
52
when they extract the logs, which provide access for the farmer. Most of this activity takes place
on land that is being cleared along new roads. Principally along the new road to Cochabamba in
provinces of Sara and Ichillo.
Timber Prices
A survey of timber prices (Cordova 1997) reports stumpage prices for 24 species in various
parts of the Chiquitania and Guaragyos of Santa Cruz, plus the integrated area of north Santa Cruz
and the river port, Puerto Villaroel en the Chapare of Cochabamba. All but five species range in
price from 7 to $ 12 per tree. Five species reached higher levels: Almendrillo of 17, Verdolago at
19, Cedro at 23, Roble at 29 and Mahogany was quoted at $ 46 per tree. There are about 2.1 cubic
meters of stumpage in each tree3.
Forest management costs
Sustainable management for a private landowner who will sell standing timber involves
maintaining control to avoid encroachment, planning for use of the resource, supervision of
extraction, and under Bolivian Law, establishment of plots for permanent measurement of tree
regeneration and growth.
The data presented here are based on estimates by BOLFOR, mostly from calculations based
on time requirements and prevailing salary levels, as little data is available from actual experiences
13 Based on an estimate obtained in Lic. Fernando Aguilar, BOLFOR.
53
when they extract the logs, which provide access for the farmer. Most of this activity takes place
on land that is being cleared along new roads. Principally along the new road to Cochabamba in
provinces of Sara and Ichillo.
Timber Prices
A survey of timber prices (Cordova 1997) reports stumpage prices for 24 species in various
parts of the Chiquitania and Guaragyos of Santa Cruz, plus the integrated area of north Santa Cruz
and the river port, Puerto Villaroel en the Chapare of Cochabamba. All but five species range in
price from 7 to $ 12 per tree. Five species reached higher levels: Almendrillo of 17, Verdolago at
19, Cedro at 23, Roble at 29 and Mahogany was quoted at $ 46 per tree. There are about 2.1 cubic
meters of stumpage in each tree3.
Forest management costs
Sustainable management for a private landowner who will sell standing timber involves
maintaining control to avoid encroachment, planning for use of the resource, supervision of
extraction, and under Bolivian Law, establishment of plots for permanent measurement of tree
regeneration and growth.
The data presented here are based on estimates by BOLFOR, mostly from calculations based
on time requirements and prevailing salary levels, as little data is available from actual experiences
13 Based on an estimate obtained in Lic. Fernando Aguilar, BOLFOR.
53
of landowners.14 The table below summarizes the fixed and variable costs. Initial costs are then
depreciated over a five-year period. Total costs are then averaged over the area of the entire forest
production unit. The costs of sustainable management practices is much higher per hectare for
smaller units-- $6.55 per hectare for a 1,000 hectare forest unit versus just over $2.73 per hectare for
a 5,000 hectare unit.
Table 12. Forest management costs: 1,000 and 5,000 hectares
1000 hectares 5000 hectares
Costs Fixed Variable Fixed Variable
Planning
Preparation of management plan and inventory 2,510 3,963
Preparation of annual operating plan and census 2,250 4,450
Production
Road building 1,800 3,600
Supervision 2,691 5,336
Forest tax 50 250
Other
Marketing 700 2,100
Total Costs 4,310 5,691 7,563 12,136
Depreciation of fixedcosts 862 1,513
Average annual cost per hectare 6.55 2.73
Source: BOLFOR
Management plans, inventory, and operating plans
As mentioned above new legislation has tightened the use of the forest resource. All forest
operations, whether a concession or a private land-owner, must now prepare a management plan and
14 Thanks especially to Lic. Fernando Aguilar, forest economist at BOLFOR, who obtained part of
the information related to costs of inventories from Dr. Erhard Dauber.
54
inventory. The cost of these plans could become an area of concern for small land-owners who
intend to incorporate forestry into their land use plans. Although there are no economies of scale
which would give advantage to larger land users, the costs, at $2,510 for 1,000 hectares and $3,963
for 5,000 hecatres, are high. Indeed, a land-owner with 1,000 hectares who wished to be involved
in forestry would have an initial outlay of over $10,000.
The operating plans are prepared annually and involve a commercial census, material
purchase and the preparation of the plan. Table 13 shows the costs of preparing annual operating
plans to be $2,250 for 1,000 hecatres and $4,450 for 5,000 hectares.
Table 12. Preparing a management plan and inventory
1000 has 5000 has
Costs of general management plan ($us)
Forest inventory (see below) 1,820 2,713
Forestry specialist 490 1,050
Preparation of document 200 200
Total 2,510 3,963
Detail of forest inventory costs:
Fotointerpretation and mapping 500 500
Planning 70 140
Field work 355 1078
Supervision 85 185
Data processing 450 450
Preparation and publication of plan 360 360
Total 1,820 2,713
Source: BOLFOR
Table 13. Costs of preparing annual operating plans
Activity 1000 has 5000 has
($us)
Commercial census 550 2,750
Preparation of Annual Operating Plan 1,400 1,400
Materials 300 300
Total 2,250 4,450
Source: BOLFOR
Timber transportation costs
Logs are usually transported over relatively short distances to sawmills. This is especially
true in the case of lumber companies that can locate sawmills in or near concessions, allowing
transport of logs within only abour 30 kilometers. General surveys of loggers and sawmill operators
(BOLFOR) determined an average distance of 52 kilometers in 19975. In a few cases, logs are
transported much larger distances. Transport costs per kilometer decrease with distance, ranging
from less than $0.10 for distances of over 100 kilometers to over $0.20 for distances under 50
kilometers. The total cost of transporting a cubic meter of log 60 kilometers is estimated at
$8.96/m3, while transporting a cubic meter 90 kilometers would cost $10.55/m3. The density, and
thus weight, of woods also has an impact on transport, with heavy woods costing perhaps 20 percent
more per cubic meter than relatively less dense medium woods.
Lumber is transported from sawmills to local market and processing centers--mainly Santa
Cruz, La Paz and Cochabamba--over distances between 300 and 1400 kilometers. Here processing
includes drying, precision cutting, and other manufacturing, such as for veneer and plywood.
15 C6rdova, BOLFOR.
Transport is usually by truck, and from more remote regions in the northwest, by riverboat. Road
infrastructure even from sawmills to major cities in most areas consists mostly of dirt roads useable
only during dry weather. Average cost of transport by truck is close to $0.065 per kilometer for a
cubic meter of lumber while the cost by riverboat is close to $0.028. The cost of transportation from
Santa Cruz to the port at Arica is $74 per metric ton (CADEX) which translates into more than $59
per cubic meter or $0.14 per board foot'6.
Rail is rarely used for internal shipment of lumber, as the cost per cubic meter kilometer is
similar to that of trucking, but implies additional handling and delivery costs on both ends.
16 Cdmara Forestal de Bolivia export data for Santa Cruz (1997) show an average weight per cubic
meter of 795 kilograms or 1.25 cubic meters per metric ton. There are about 424 board feet in a
cubic meter.
Accounting for regional productivity, forest management costs, and land use taxes in the
analysis of land conversion from forest to pasture in the Bolivian Lowlands
by
F. D. Merry'*, P. E. Hildebrand', D. R. Carter', and P. Pattie2
1 University of Florida, Gainesville, Florida, USA
2 Chemonics International, Washington, D.C, USA
Manuscript for submission to
LAND USE POLICY
March, 2001
*Corresponding Author
528 SE 1st Avenue
Gainesville, Florida, 32601
email: fdme@gnv.ifas.ufl.edu
Accounting for regional productivity, forest management costs, and land use taxes in the
analysis of land conversion from forest to pasture in the Bolivian Lowlands
Abstract
Current forest legislation in Bolivia decrees that all land in forestry be assessed a uniform
area tax. This research suggests that land taxes should vary according to regional productivity and
forest potential. Five regions, spread throughout the Lowlands, are identified and their likelihood
for land conversion from forest to pasture for cattle ranching is estimated. The variability of
conversion rates showed that marginal forest areas are subject to conversion to other, more
productive uses, and thus could use a reduction in the area-based land use tax to compete. Possible
land conversion rates were estimated at 1.3 % per year for the region described as Guarayos, and 1.1
% per year for the Chiquitania region both well above the national deforestation rate of 0.50 %.
The conversion estimate for the regions described as Andean Foothills, Paragui and Pando, suffered
no conversion from forest to pasture. This is a result of both high forest productivity and poor cattle
production conditions. Also we show that forest management costs and land use taxes affect land
conversion to pasture; and that increasing pasture taxes, rather than changes in the forest land tax,
would be an effective means of curtailing land conversion on areas with low forest management
costs. In areas with high forest management costs land use taxes were less effective in manipulation
of the land conversion rates.
Key Words: Linear programming, land use policy, cattle, forestry.
1. Introduction
Enactment of new forest legislation in 19961 in Bolivia, together with a new agricultural law2,
has had the dramatic effect of reducing the area under forest concessions from 20 million hectares
to about 6 million. But, whereas forestry activity could previously be undertaken only on forest
concessions, the new legislation allows for forest management on private lands. Productive use of
the forest, however, on both private and concession lands, now requires formal management
planning and is subject to a uniform annual tax per hectare3.
The practical side of supervising the forest sector makes ostensibly simple taxation methods,
such as uniform area or value taxes, more appealing (Hyde and Sedjo 1992). The uniform taxes
appear to require less information on the nature and productivity of the forest, but it is this very
information that allows other, more complex, forms to accurately assess the tax, and in the absence
of which, the uniform taxes can be inaccurate (Vincent 1990).
Most of the former concessions, now returned to the state, were located in land-use zones
designated for combined cattle and forestry production. It follows that a private landowner who gains
control of that land will have either sustainable forestry or cattle production as the two primary land-
use options. What is not clear, however, is the combination of forest and pasture that will maximize
the land owners benefits, how regional productivity will affect the land use choices, and whether
variations in land taxes or planning requirements can be effective in encouraging forestry rather than
pasture.
1 Nueva Ley Forestal, Ley Numero 1700, del 12 de Julio de 1996.
2 The Ley de Servicio National de Reforma Agraria (Ley No. 1715 known as the ley INRA)
3 On private lands, however, the tax is on the area harvested rather than the total area.
3
The objective of this paper is to examine the new land use dynamics in the Lowlands created
by the removal of forest concessions and the imposition of new land use taxes and land use planning
requirements. To do this we develop and empirical model to assess the rate of land conversion
between these two important land use options in the Lowlands. We ask how land conversion will
change when subject to regional productivity differences, to changes in forest management costs
(including land use planning requirements), and to variations in land use taxes.
The simulation results clearly indicate that two regions of the Lowlands, the Chiquitania and
Guarayos, are higher risk areas, and will likely suffer the most conversion from forest to pasture.
All of the other regions the Andean Foothills, Paragud, and Pando appear, under current
conditions, to be in little risk of major land conversion within the next 20 years; due both to high
forest productivity as well as poor cattle production conditions. In addition, the results suggest in
areas where cattle is competitive with forestry, higher forest management costs lead to higher rates
of land conversion to pasture; and the impact of changes in land taxes (i.e. changes in pasture or
forest land taxes) is more drastic on areas with lower forest management costs. The results from this
research shows that there are two regions susceptible to land conversion and that flexibility in
forestry tax rates, or tax rates on other land uses, would encourage sustainable forest production in
regions of higher risk of land conversion.
The rest of the paper is as follows: a brief discussion of the current land taxes and land use
planning requirements; an introduction to forestry and cattle production in the Lowlands; a
description of the model used for the empirical analysis, and an overview of the five regions
identified throughout the Lowlands that will be used to compare regional susceptibility. And finally,
a review of the results and a conclusion.
2. Bolivian land taxes and land use planning requirements
Under the new agricultural and forestry laws new regimes have been established for: taxes
and land use planning in forestry; land conversion from forest to other land uses; and land use taxes
in pasture land. For forested lands the new Forest Law assesses a per hectare uniform area-based
forest land use tax. This tax is currently set at $1.00 and can be adjusted upward, but can never fall
below the $1.00 value. From the description, the tax is uniform across all ranges of productivity and
profitability of forest lands; no compensation is made, for example, for forests with low stand
densities, a preponderance of lesser valued species, or for forests in regions with high transportation
costs. On private lands, this tax is assessed annually on the area of harvest. On concessions,
however, it is assessed on annually for every hectare regardless of harvest volume or area.
The current law also requires preparation of "Land Use Plans" for privately-owned forest
properties. Individual land use plans are similar to the requirements faced by the forest
concessionaires. Individual land use plans require the classification of areas for intensive annual
cropping, perennial crops, pastures, open grazing, and forestry. The plans also prescribe minimum
conservation practices. Individual property land use plans are to be registered with land titles and
become legal zoning restrictions or easements applying to current and future owners. Just as with
forest management plans for concessionaires, individual property land use plans are prepared by
professionals who must be hired by the landowners. These land use plans are expensive and have
a definite scale bias against the smaller land holders. Table 1 shows the per hectare costs of
preparing a forest inventory and management plan where the smallest holding of 500 hectares pays
a startling $30.93 per hectare and the largest example of 200,000 hectares pays only $0.11 per
hectare.
The costs associated with sustainable forest management, which include a forest inventory
and management plan were further detailed by the Bolivia Sustainable Forestry Project (BOLFOR),
and are estimated at $6.50 per hectare for 1,000 hectares and $2.68 per hectare for 5,000 hectares
(Table 2) The planning and management process incurs both initial and annual costs. Since these
plans are only good for five years, after which time they must be renewed, initial costs are
depreciated over this period. In this estimate, the two of the major costs are the preparation of the
management plan and inventory and the preparation of the annual operating plan and census. These
costs show marked differences in the per hectare costs between 1,000 and 5,000 hectares.
Management plans and inventory are estimated to cost $2.51 per hectare for and area of 1,000
hectares and $0.79 per hectare for an area of 5,000 hectares4. Similarly, the cost of preparing annual
operating plans and the harvest area census are estimated at $2.25 per hectare for 1,000 hectares and
$0.89 for 5,000 hectares.
For land conversion from forest to any other use, the presentation and approval of an
individual property land use plan and land clearing plan is required. Any timber sold as a result of
land clearing requires a "certificate of forest origin" and payment of an area and a volume-based tax.
The area tax for clearing land is currently set at 15 times the per hectare tax paid by forest
concessions, or $15/hectare. The volume tax is computed at 30 percent of the value of timber at the
land clearing site, according to a price list established by the Superintendencia Forestal. The
landowner pays 15 percent of the value and the buyer pays the other 15 percent. From a practical
4 These estimates are slightly lower than the estimates in Table 3, but show similar economies of scale.
6
standpoint, if the landowner finds the value taxes overly burdensome the timber will be burnt off,
which is, indeed, common practice.
Land converted to pasture incurs a progressive tax based on the total land value declared by
the owner; the scale begins at 0.35% and increases to 1.50% of the value without improvements
(Table 3). Actual land prices for cleared pasture land may be as high as $500 per hectare in
accessible areas along major roads in regions such as the Chiquitania of Santa Cruz. Outside these
zones, however, prices decrease dramatically to $20 or lower. Generally, pasture lands with limited
access may be valued between $100 and $250 per hectare. Depending on the size of ranches and the
total value, taxes on land cleared and planted to pasture are typically between $0.20 and $0.50 per
hectare (Pattie and Merry, 1999).
3. Forestry and cattle production in the Lowlands
Approximately 51 percent of Bolivia's 1,100,000 Km2 is covered in forest (Lopez 1993).
The Department of Santa Cruz has long been the most important Department in the forest sector,
recording an average of 60 percent of official total log production and 80 percent of mahogany
exports during the period 1980-1993, and accounting for a total export value of 44.5 million dollars
in 1994 (CNF 1995). Of a total 36.5 million hectares in the Department, approximately 18 million
(49 percent), are described as having regular, high, or very high forestry potential.
The Bolivian forest sector, however, is in the midst of change. Nineteen new concessions
have been allocated in the far northern Department of Pando. And, with mahogany stocks in the
decline (Stolz and Quevedo 1993; Merry and Carter 1999), there has been a concerted effort to
change from selective harvest techniques to sustainable forestry management that relies on lesser-
7
known (and lesser-valued) species. Indeed, Cedro (Cedrelafisilis) has become the principal export
wood in terms of both volume and value (Table 4). Given the new land use regulations (see section
2), forest land now under the domain of the SF must be managed in a sustainable fashion, and
therefore, in an effort to examine the possible impact of the new forest legislation on land use, this
paper compares sustainable forest production (including forest management planning and tax costs)
to the most likely alternative use cattle production.
Except for the area immediately surrounding the City of Santa Cruz, cattle production
throughout the Bolivian Lowlands is an extensive operation. The current Bolivian cattle population
is approximately six million head, an increase of nine percent from the 1990 population. Of the six
million head, 2.7 million are to be found on the vast natural grasslands of the Beni and 1.5 million
in Santa Cruz, such that these two Departments alone comprise 70 percent of the total herd. Many
areas of the Bolivian Lowlands, including those once in concession, are suitable for cattle production
or a combination of forestry and cattle production (PLUS 1995). For example, in the department of
Santa Cruz, 64 percent of the land is deemed suitable for cattle production in one form or another.
4. A linear programming model to analyze land conversion from forest to pasture
Linear programming is a general optimization technique that can be used to examine the
optimal allocation of scarce resources (Buongiomo and Gilless 1987). The methodology has been
used successfully for many years in operations research for both agricultural and forestry production
(e.g., Howard 1993, Nicholson et al. 1994, Osho 1995), and in the empirical estimation of
deforestation at the household and firm level (Kaimowitz and Angelsen 1998).
The model presented here is a multi year dynamic optimization model that examines ajoint
cattle and timber production scenario, and evaluates how selected parameters will affect the rate of
land conversion from timber to pasture. The model has a twenty-year time horizon, separated into
ten periods of two years each. The objective function maximizes the net present value of terminal
year capital assets and bi-annual disposable income. Capital assets are described by land, both in
forest and pasture, and the cattle herd. Disposable income is the cash that the landowner can remove
from the system every two years during the 20-year horizon. The mathematical structure of the
model is presented below:
Mathematical structure and description
T
Max ( + r)-W + (1 + r) TVHT+(1 + r)- VFT+(1 + r)- TVPT
s=l
s.t.
CSHt CSHt -TI SB4 SH2t + NCC + PCBt + MC, + PT + DI = 0
t
F + l Pt 3,000
s= 1
-DIt + Wt 0
TIt (NRF)Ft = 0
t
TH- /SR X t O
s=l 1
THt TC BL B2 B4t H2 = 0
TCt + (1- CR)TCt BCt HF2Kt = 0
BLt (fB)TCt = 0
B2t (a)TCt =0
B4t -(1- DL)B21t 1= 0
SB4t B4t = 0
HF2t (a)TCt = 0
HF2 SH2t HF2Kt = 0
t
PTt (8) Pt = 0
s= 1
THT + VHt : 0
FT +VFT 0
T
E P + VPT 0
s=1
VH, VF, VP, F, P, TI, TH, SR, TC,
BL, B2,B4, H2, HF2K, PCB, CSH, DI, W 0
Where: t is the period identifier beginning at 1; T is the last year of the model; r is the
discount rate; VH is the terminal value of the cattle herd; VF is the terminal value of the forest land;
VP is the terminal value of the pasture; F is the hectares in forest; P is the hectares converted to
pasture; TI is the timber income; NRF is the net returns per hectare of forest; TH is the total cattle
herd; SR is the stocking rate; TC is the total cows; BL is the number of service bulls; B2 is the
number of young bulls; B4 is the number of bulls for sale; H2 is the number of heifers; CR is the
culling rate for cows (inculdes death loss); BC is the cows purchased; DL is the death loss per
period; SB4 is the bulls sold; SH2 is the heifers sold; HF2K is the heifers kept for the cow herd;
CSH is the cash available in any period; TI is the timber income; NCC is the net cost of land
conversion; PCB is the cost of buying cows and bulls; MC is the herd maintenance cost; PT is the
pasture tax; DI is the disposable income; W is the disposable income withdrawn per period; a is the
calving rate of 60 percent; p is the breeding bull to cow ration; and 6 is the pasture tax rate.
For the simulations of regional susceptibility an area of 3,000 hectares is initially allocated
to the landowner. In the subsequent analysis of forest management costs and land use taxes the
initial land allocations are 1,000 hectares and 5,000 hectares. The land is free, and assumed to be
of a single quality class and completely forested. The cost of sustainable forest management for
1,000 and 5,000 hectares was estimated by the Bolivian Sustainable Forest Management Project
(BOLFOR-USAID) and are presented above in Table 2. These values were pro-rated for the area
of 3,000 hectares.
Each hectare of land earns bi-annual revenue under a sustainable forest management
program. Stumpage volumes were collected from management plans of regional concessions, and
are subject to universal 4 percent annual allowable cut. A net return per hectare of forest is
11
calculated by subtracting the management costs from the gross value of the allowable cut. The
stumpage price is initially set at $4.00 per m3. Estimates for stumpage prices throughout Bolivia
ranged from $3.53 to $5.66 (Soto 1996, Cordova 1998). Thus we also include a sensitivity analysis
where a range of stumpage values are used.. The ending value for forestland is determined as the
net discounted value of perpetual sustainable forest production into infinity, also known as land
expectation value.
In order to start a cattle herd, the landowner must first clear the land. Pasture clearing costs
were generated through interviews with land clearing companies and ranchers and range from $350
to $605, including a $15 per hectare clearing tax. There is no initial capital available. It is assumed,
however, that at the time of clearing, the landowner will sell all available commercial timber. A 30
percent tax is levied on the sale of timber from cleared lands. Thus, the net cost of land conversion
is the cost of clearing minus the value of the stumpage net of taxes. The final pasture land value
reflects returns to investment in clearing.
Finally, with standard extensive management techniques employed throughout all regions,
cattle production costs are expected to be uniform. Table 5 gives the values for variables used
throughout the analysis. Cattle production costs were found in the annual publication of agricultural
statistics of the Eastern Agricultural Association (CAO 1997). In addition, most sales occur at a few
abattoirs in major metropolitan areas and so prices for beef products are expected to be similar, at
$1.50 per kilo, across regions. The major difference between the regions is transportation, and
indeed, the cost of getting cattle to and from the ranches varies tremendously. For example, it costs
$35.18 to transport an animal to market from the Guarayos region, but $105.16 from the Pando.
Transportation costs were estimated in through informal interviews with transportation associations,
12
ranchers, and individual truckers. These transportation costs, in addition to the values for other
selected program variables, are shown in Table 6, and illustrate all of the important regional
differences.
5. Regional site descriptions and results
The site descriptions are of the five regions examined in the analysis are given below,
followed by the respective land conversion estimates. The regions stretch across the length of the
Bolivian Lowlands and cover most of the area where there is potential for conflict between cattle
production and forestry. It is important to note that the geographic references for are for the purpose
of this study only, and although they bear some resemblance to actual regions, the borders are those
shown in Figure 1. The regions are described as: (1) Chiquitania, (2) Guarayos, (3) ParaguB, (4)
Andean Foothills, and (5) Pando; the descriptions are taken, in part, from the following studies:
Killeen et al. (1993), MDSMA (1995) and ITTO (1997). In addition to export volumes and values
for all major species, Table 4 shows which species are available within each of the regions.
Chiquitania situated east of the Beni plains and comprising approximately 7 million
hectares, the Chiquitania site is on the Brazilian Shield and averages 1,000 meters above sea level
(m.a.s.l.). It is an undulating terrain with mostly subtropical dry forest that receives annual rainfall
from 900-1,200 mm. This region has a seven month dry season allowing harvest operations for six
to seven months of the year. The soils have the following characteristics: poor chemical content,
well-drained, red clay soils, acidic, high exchangeable aluminum, and low fertility (oxisols, ultisols).
There has been a permanent settlement in this area for a long period, primarily by indigenous
populations and there is a high level of cattle production activity (Killeen 1991). The forest is
comprised mainly of very dense hardwoods with no mahogany.
Guarayos the Guarayos region is located immediately northwest of Chiquitania and covers
approximately 8.5 million hectares. As part of the Brazilian Shield it shares many characteristics
with the Chiquitania. The primary difference is a higher annual precipitation of approximately 1,800
mm. It is classified as a subtropical moist forest. The soils are primarily well-drained, red clay soils,
high exchangeable aluminum, and low fertility (oxisols, ultisols). Harvest operations in this area can
take place four to five months of the year but are impeded by heavy rains during the wet season. The
Guarayos site contains forest with mixed inventories of hardwoods and softwoods. Mahogany
occurs in low volumes. There is a strong presence of cattle production in this region.
Paragud the Paragud area represents a region of approximately 5.5 million hectares. It is
a subtropical moist forest that receives an annual rainfall of 1,600 mm and is an average of 800
m.a.s.l. For the most part the soils are well-drained, red clay soils, with low fertility (oxisols,
ultisols). Located in the far northeast section of the Department of Santa Cruz, and the southeastern
section of the Department of the Beni, this remote region has poor infrastructure and very low
population pressures. Transportation costs for both agricultural and forestry commodities are very
high with log trucks often taking three to four days to reach Santa Cruz for a distance of between
500 and 800 kilometers. The dry season is prolonged and severe, allowing forest harvest for five-six
months, but creating problems for agricultural production. There is currently little cattle production
activity in the area. There are a high percentage of dense hardwoods, and a low-medium frequency
of mahogany.
Andean Foothills the area described here as the Andean Foothills stretches through the
Departments of La Paz (central) and the Beni (western); east from the Andean foothills, and west,
southwest of the natural grasslands of the Beni. The Andean Foothills region is a subtropical moist
forest, covering 5 million hectares, with an average annual rainfall of 2,000-2,800 mm. This area
is an ancient alluvial plain with sandy loam soils of variable texture and adequate fertility. Poor
drainage and high rainfall combine to create frequent flooding, saturated soils, and very difficult
production conditions for six months of the year. The timber harvest season is limited to three
months per year, but regardless, has good forestry potential. The forest is comprised mainly of white,
softer woods, with Ocho6 (Hura crepitans) being the dominant species. There are strong
colonization pressures in some areas in this site, and a moderate level of cattle production.
Pando This region lies within the new forestry frontier and consists of the far north of the
Department of La Paz and the entire Department of Pando. It has received 19 new concessions,
totaling 1.5 million hectares, in the Department of Pando alone, since the passing of the new forest
legislation (Superintendencia Forestal 1998). Precipitation ranges from 1,800-2,000 mm and there
are only three dry months. The predominant soils are again oxisols and ultisols. The forests are
considered to be more productive than in the Guarayos and ChiquitaniA sites, but higher precipitation
constrains the harvesting season duration. Cattle production options are low, however, with
considerable market and transportation constraints.
6. Results
Regional susceptibility to land conversion
The simulation results for the individual regions clearly demonstrate differences in
susceptibility to land conversion5. Table 7 shows the simulation results for land conversion rates
for the five regions. The results are for total area converted after 20 years, total percentage
conversion, and an annual rate of conversion.
The Guarayos and Chiquitania regions were the two regions most susceptible with 27 and
22 percent conversion, respectively, over the 20-year simulation timetable. This reflects average
annual conversion rates of 1.3 percent for Guarayos and 1.1 percent for the Chiquitania. In the other
three regions described by the Andean Foothills, the Pando and ParaguA there appears to be little risk
of land conversion with 0 percent converted over the 20 year period.
Impact offorest management costs
Earlier, Table 2 gave the forest management costs for an area of 1,000 acres as $6.50 per
hectare (hereafter referred to as high management costs) and $2.68 per hectare for an area of 5,000
hectares (low management costs). The model was run, with average cattle production inputs, to
compare these two forest management cost scenarios on both land areas.. The results show that we
can expect 5 percent land conversion to pasture on a 1,000 hectare area with high management costs,
but, when run with low management costs, we can expect only 2 percent conversion. On the 5,000
hectares, we can expect 28 percent land conversion with high management cost, and 11 percent with
low management costs. This shows that forest management costs which include land use planning
requirements can drastically affect the level of land conversion, and, indeed, appear to have a greater
relative impact on larger areas.
5 The current national average annual rate of deforestation for Bolivia is 0.5 percent (Pacheco 1998, but This
includes conversion to all other land uses, not solely to pasture.
16
Impact ofrelative changes in land taxes
The model was run with the forest tax held constant at its current rate of $1.00 per hectare,
and the pasture tax varied increased from its estimated rate of $0.50 to a rate of $2.00 per hectare for
both small and larger areas (1,000 and 5,000 has), with their respective forest management costs
(Figure 2). On the larger area, the results showed a steady decline in land conversion to as the tax
on pasture land increases, until there is no conversion at $2.00. On the 1,000 hectares, however,
increasing the pasture tax had little effect on the overall rate of land conversion. Results in Table
8 further develop this idea for the area of 5,000 hectare; in this case, the pasture tax is held constant
(at $0.50 per hectare) and the forest land tax is varied from $0.00 to $2.00 per hectare. This results
shows the land conversion rate unchanged. These results suggest two points (1) that under
conditions of high forest management costs, relative land taxes are not a determinant of land
conversion; and (2) under conditions of low forest management costs, changes in pasture taxes are
more effective determinants of land conversion than forest taxes.
7. Conclusion
The recent release of land from forest concession to private ownership stands to increase the
availability, and incentive, of land to be converted from forestry to alternative land uses in the
Lowlands of Bolivia. The monitoring and control of forest production on both private and public
land is in the hands of the Superintendencia Forestal and it is important that this agency be able to
efficiently allocate its limited resource to best support sustainable forest production throughout the
Lowlands. In public forest lands concessions the task is relatively simple; ensure the compliance
with prescribed best forest management practices. On private lands, however, sustainable forest
17
management is complicated by the option to convert to other, more profitable, land uses. The SF
must then manage these lands with the heightened concern over the potential for land conversion.
This paper inquires into the potential interaction of two important land use options in the
Lowlands sustainable forestry and cattle production under the new economic parameters of the
land use legislation adopted by Bolivia in 1996/97. We show that regional variation in productivity,
forest management costs, and land use taxes all affect the likelihood of land conversion. The results
show that land taxes and land use planning requirements which increase marginal costs of forestry
need careful consideration before broad application.
The simulation results clearly indicate that the two regions of the Brazilian shield,
Chiquitania and Guarayos are higher risk areas, and will likely suffer the most conversion f rom
forest to pasture. All of the other regions the Andean Foothills, ParaguA, and Pando appear, under
current conditions, to be in little risk of major land conversion within the next 20 years; due both to
high forest productivity as well as poor cattle production conditions. The results also imply that
forest management costs which include land use planning requirements can drastically affect the
level of land conversion, and, indeed, appear to have a greater relative impact on larger areas.
Finally, these results suggest two points (1) that under conditions of high forest management costs,
relative land taxes are not an important determinant of land conversion; and (2) under conditions of
low forest management costs, changes in pasture taxes are more effective determinants of land
conversion rates than forest taxes.
The policy implications of this work lie in the freedom to vary land taxes and land use
requirements to minimize land conversion.. At present forest land taxes in Bolivia are set at a
national level. This work shows that increased flexibility would encourage sustainable forest
production in marginal regions and thus reduce the risk of land conversion to other uses.
Literature cited
Buongiorno, J. and J. Gilless. 1987. Forest Management and Economics. Macmillian
Publishing Company, New York NY. 285pp.
Camara Agropecuaria del Oriente (CAO). 1997. Numeros de Nuestra Tierra, Santa Cruz, Bolivia.
Camara Nacional Forestal (CNF). 1995. Estadisticas de Approvechamiento, Exportacion y
Commercializacion de Productos Forestales. Santa Cruz, Bolivia.
C6rdova, Javier. 1998. Encuesta de Precios de Venta de Madera en Diferentes Etapas de
Utilizaci6n. Pasantia. Universidad Aut6noma Gabriel Rend Moreno. Santa Cruz.
Howard, A. F. 1993. A linear programming model for predicting the sustainable yield of timber
from a community forest on the Osa Peninsula of Costa Rica. Forest Ecology and
Management, 61:29-43.
ITTO. 1996. The Promotion of Sustainable Forest Development in Bolivia. Report Submitted to
the International Tropical Timber Council by the Mission Established Pursuant to Decision
2(XVII).
Kaimowitz, D. and A. Angelsen. 1998. Economic Models of Tropical Deforestation A Review.
CIFOR Special Publication, Jakarta, Indonesia. 139 pp.
Killeen, T. J. 1991. Range management and land-use practices in Chiquitania, Santa Cruz, Bolivia.
Rangelands 13:59-63.
Killeen et al. 1983. Guia de Arboles de Bolivia. Herbario Nacional de Bolivia and Missouri
Botanical Garden.
Lopez, J. 1993. Recursos Forestales de Bolivia y su Aprovechamiento. Cooperaci6n Tecnica
Holandesa, Artes Graphicas, La Paz.
Merry, F. D. and D. R. Carter. 2001. Factors affecting Bolivian mahogany exports with policy
implications for the forest sector. In Press Forest Policy and Economics
Ministerio de Desarollo Sostenible y Medio Ambiente. 1995. Mapa Forestal de Bolivia, Memoria
Explicativa.
Nicholson, C. F., D. R. Lee, and R. N. Boisvert. 1994. An optimization model of the dual-purpose
cattle production system in the humid lowlands of Venezuela. Agricultural Systems 46:311-
334.
Osho, J. S. A. 1995. Optimal sustainable harvest models for a Nigerian tropical rainforest. Journal
of Environmental Management 45:101-108.
Pacheco, P. 1998. Estilos de desarrollo, deforestaci6n y degradaci6n de los bosques en las tierras
de Bolivia. CID, La Paz, Bolivia. 389 pgs.
Pattie P. and F. D. Merry. 1999. Forests versus Livestock: An Economic Assessment of Choices
Facing Landowners in the Bolivian Lowlands. Technical Document # 74/1999, Bolivian
Sustainable Forestry Project (BOLFOR), USAID, Santa Cruz, Bolivia.
Soto, Ademar Juan Carlos. 1996. Encuesta de Precios de Venta de Madera en Diferentes Etapas de
Utilizaci6n. Pasantia. Universidad Aut6noma Gabriel Ren6 Moreno. Santa Cruz, Bolivia.
Stolz, R., and L. Quevedo. 1992. Estudio del Sector Forestal del Departamento de Santa Cruz.
Proyecto de Protection de los Recursos Naturales en el Departamento de Santa Cruz. Santa
Cruz, Bolivia.
Figure 1. Geographical site descriptions for five regions in the Bolivian Lowlands
SANTA CRUZ
1. CHIQUITANIA
C-9
1 SAN JOAQOIN
2 CIUAGRO
3 LA LUNA LTDA.
A YUCUIMO
5 TARUIA LTDA.
6 LACH1ONTA SRL
7 LOMERIO rlndigenaunalarrkay(
8 LAMINADORA SAN MIGUEL
9 Srs BOLIViA
BENI
POTOSI
Figure 2. The effect of pasture taxes on conversion under conditions of high and low forest
management costs
20-
5,000 hectares
S15- Low management costs
10-
1,000 hectares
High management costs
0
N
g -<--------*----- --
&."s
$0.50 $1.00 $1.50 $2.00
Pasture tax ($/hectare)
Table 1. Economies of scale in forest inventory preparation ($US, 1995)
Area Photo & Planning Fieldwork Supervision Data Document Total cost Cost per
(has) map &transport processing preparation of inventory hectare
100 100 120 325 1,000 48 1,500 3,093 30.93
500 100 120 650 1,000 60 1,500 3,430 6.86
1,000 100 120 975 1,000 95 1,500 3,790 3.79
2,000 100 120 1,300 1,000 144 1,500 4,164 2.08
5,000 100 180 2,080 1,600 220 1,500 5,680 1.14
10,000 250 240 2,925 2,000 250 1,500 7,165 0.72
15,000 330 270 3,575 2,200 257 1,500 8,132 0.54
20,000 400 300 4,095 2,400 266 1,500 8,961 0.49
25,000 480 330 4,615 2,600 280 1,500 9,805 0.39
30,000 550 360 5,005 2,800 296 1,500 10,511 0.35
50,000 850 420 6,500 3,000 370 1,500 12,640 0.25
100,000 1,600 480 9,165 3,400 450 1,500 16,595 0.17
200,000 3,100 600 12,935 4,000 550 1,500 22,685 0.11
Source: Bolivian Sustainable Forest Management Project (BOLFOR) published as "Normas tecnicas para la elaboracion de
instruments de manejo forestal". Resolucion Ministerial No. 62/97, La Paz, Bolivia (1997:V-46).
Table 2. Costs of sustainable management on private forests of 1,000 and 5,000 hectares.
Activity 1,000 hectares 5,000 hectares
Initial Yearly Initial Yearly
($US) ($US) ($US) ($US)
Planning
Preparation of management plan and inventory 2,510' 3,963a
Preparation of annual operating plan and census 2,250 4,450
Production
Road Building 1,800 3,600
Supervision 2,691 5,336
Marketing 700 2,100
Total Costs 4,310 5,641 7,563 11,886
Depreciation of Initial Costs 862 1,513
Average Annual Cost Per Hectare 6.50b 2.68b
Source: Bolivian Sustainable Forest Management Project (BOLFOR).
SThe values used here for management plans and inventory are lower than those presented in Table 3.
b Forest land use taxes not included
Table 3. Land tax schedule for livestock ranches, 1996.
Declared total value Fixed tax Plus variable rate On amount
($US) ($US) (%) over ($US)
0 to 40,800 0 0.35 0
40,800 to 82,000 143 0.50 40,800
82,000 to 122,500 347 1.00 82,00
> 122,500 756 1.50 122,500
Source: Pattie and Merry (1999)
Table 4. Selected wood exports from Bolivia", 1998, with regional species distribution
Vol. ( m) Value ($US) CH
36,916 21,207,031 X
21,914 18,020,299
Species Scientific Name
Cedro Cedrelafisilis
Mara Sweitenia macrophylla
(mahogany)
Roble Amburana cearensis
Yesquero Cariniana estrellensis
Lenga Nothofagus pumilio
Palo Maria Calophyllum brasilense
Mara Macho Terminalia amazonia
Ocho6 Hura crepitans
Morado Machaerium scleroxylon
Bibosi Ficus sp.
CambarA Erisma calcaratum
Paqui6 Hymenaea coubaril
Tarara Centrolobium microchaete
Sirari Ormosia coarctata
Tajibo Tabebuia heptaphylla
Almendrillo Dipteryx alata
Quina Cinchona calisaya
Plumero Vochysia lanceolata
Cuta Phyllostylon rhamnoides
Ajunao Pterogyne nitens
Soto Schinopsis brasiliensis
Picana Negra Cordia alliodora
Quina Quina Cinchona officinalis
Jichituriqui Aspidosperma
cylindrocarpon
Curupai Anandenanthera colubrina
14,580,969 X
3,604,857 X
2,385,986
2,309,137
2,278,468
2,258,255
1,414,498 X
1,365,653
936,016
441,109 X
428,781 X
347,764 X
220,405 X
176,956
157,239
141,355
138,411 X
104,127 X
81,500 X
66,167 X
46,021
45,706 X
48 45.122 X
GU PG
X X
X X
AF PA
X X
X X
X
X X X X
X X X
X X
X X X X
X X X X
X X
X
X X
X
X X
X X
X X
21,037
5,279
232
2,735
1,440
2,472
1,079
1,379
1,951
418
575
490
483
153
465
346
120
204
283
82
36
145
Source: Ministerio de Agricultura, Ganaderia y Desarollo Rural. 1998.
a Includes value added products; sawnwood, however, accounts for 75 percent of export volume and 60 percent of
export value. Total volume of exports is 108,800m3 with a total value of approximately $76 million.
X X X X
X X
X X
X X
X
X X X X
X X
X
X
Table 5. Base levels for model variables
Description value unit Description value unit
CATTLE GENERAL
Purchase price cow 250.00 $/head Initial capital 0.00 Dollars
Purchase/sales price bull 280.00 $/head Discount rate 5.00 Percent
Sales price heifer 200.00 $/head Tax on land clearing 15.00 $/hectare
Annual maintenance cost 12.60 $/head Pasture carrying capacity 1.5 ha/head
Calving rate 60.00 Percent
Culling rate 10.00 Percent
Annual herd loss 2.00 Percent
Table 6. Regional variation for selected variables
Variables Unit Chiquitania Guarayos Pando Andean Paraguf
Foothills
Total stumpage volume m3 40.00 65.00 73.00 75.00 83.00
Annual allowable cut m3 1.60 2.60 2.92 3.00 3.32
Stumpage value $/m3 4.00 4.00 4.00 4.00 4.00
Net returns per hectare forestry $/ha 1.76 5.76 7.04 7.36 8.64
Value forest land $/ha 35.20 115.20 140.80 147.20 172.80
Pasture establishment $/ha 350.00 495.00 550.00 555.00 605.00
Net stumpage revenue $/ha 112.00 182.00 204.00 210.00 232.00
Value pasture land $/ha 238.00 313.00 345.00 345.00 373.00
Cattle transport costs $/head 39.40 35.18 105.16 55.66 70.68
Table 7. Regional comparison of land conversion to pasture (3,000 hectare area)
Result Guarayos Chiquitania Andean Paragub Pando
Foothills
Hectares converted(20 yrs) 804 651 0 0 0
Percent conversion (20 yrs) 27 % 22 % 0% 0% 0%
Annual rate 1.3 % 1.1 % 0% 0% 0%
a. The average annual rate of deforestation for Bolivia is approximately 0.50 % (Pacheco 1998).
Table 8. Variations in forest and pasture land taxes: their effect on land conversion
and government tax earnings (5,000 hectares).
Forest land tax
($/ha/yr)
$0.00
$0.50
$1.00
$1.50
$2.00
$1.00
$1.00
$1.00
$1.00
$1.00
Pasture land tax
($/ha/yr)
variation in forestry tax
$0.50
$0.50
$0.50
$0.50
$0.50
variation in pasture tax
$0.00
$0.50
$1.00
$1.50
$2.00
Land conversion
(%)
11%
11%
11%
11%
11%
17%
11%
7%
6%
0%
Can increasing net returns per hectare in sustainable forestry exacerbate
land conversion to pasture? A case study in Santa Cruz, Bolivia.
1. Introduction
Enactment of new forest legislation in 1996' in Bolivia, together with a new agricultural law2,
has had the dramatic effect of reducing the area under forest concessions from 20 million hectares
to about 6 million. But, whereas forestry activity could previously be undertaken only on forest
concessions, the new legislation allows for forest management on private lands. Productive use of
the forest, however, on both private and concession lands, now requires formal management
planning and is subject to a uniform annual tax per hectare3.
Most of the former concessions, now returned to the state, were located in land-use zones
designated for combined cattle and forestry production. It follows that a private landowner who
gains control of that land will have either sustainable forestry or cattle production as the two primary
land-use options. What is not clear, however, is whether individual private landowners will find
it within their interest to assume responsibility for sustainable management of the forests, or whether
the land will be cleared to provide pasture for cattle. This is a new question in the old debate on land
use and, in particular, between forests and pasture land.
For the past two decades, there has been extensive debate on the interaction between beef
production and tropical forestry in Latin America, the primary concern of which has been
deforestation to provide land for pasture. The increase of land allocated to cattle production at the
expense of natural forests is said to be the result of several factors such as land tenure policies,
extensive beef production systems, low timber values, production characteristics of cattle, markets
1 Nueva Ley Forestal, Ley Numero 1700, del 12 de Julio de 1996.
2 The Ley de Servicio National de Reforma Agraria (Ley No. 1715 known as the ley INRA)
3 On private lands, however, the tax is on the area harvested rather than the total area.
1
and prices favorable to beef production, government subsidies for livestock credit, and road
construction (Meyers 1981; Nations and Komer 1983; Mahar 1988; Binswanger 1991; Hecht 1992;
Serrao and Toledo 1992, 1993; Kishor and Constantino 1993; Kaimowitz 1996). The literature has
focused, with good reason, on two geographic regions: Central America, where the concern was that
forests were being destroyed to produce cheap beef for the US market ( DeWalt 1982; Endelman
1985; Browder 1988; Graciela 1989), and Brazil, where vast areas of the Amazon were cleared for
as a result of government subsidies to expansion (Hecht et al. 1988; Moran 1993; Barbier et al.
1994; Mattos and Uhl 1994; Arima and Uhl 1997; Faminow 1997). Although Central America and
Brazil have received much of the attention, the interaction between tropical forestry and cattle
production is apparent in many countries of Latin America, including Bolivia. Indeed, cattle
production has been part of the land use mosaic in the Bolivian lowlands for centuries and will
undoubtedly remain so for the foreseeable future(Killeen 1991). The passage of new forest
legislation in 1996, however, has changed the face of the forest industry and in doing so may also
affect the cattle industry, and the potential for land conversion.
In this paper we develop a model to examine land conversion from sustainable forestry, as
defined under the new forest legislation, into pasture for cattle production. On a homogenous -
single quality class land area we address how conversion will change when subject to changes in
net returns to sustainable forestry. The results suggest that in many circumstances, income from
forestry releases a capital constraint, and that, to a point, increasing per hectare returns to sustainable
forestry can serve to exacerbate land conversion to pasture. The rest of the paper is organized as
follows: First, a brief review of the economics of land conversion, then a description of the forestry
and cattle sectors of the Lowlands, a presentation of the model, and finally a discussion of the results
2
The economics of land conversion
The owner of forest land who is considering the option of cattle production on all or part of
his land faces a difficult decision of when and how many acres to convert from forest to pasture.
Due to the complex ecosystem of tropical forests, and the time that it takes for these natural forests
to regenerate, this decision is essentially irreversible. This decision is rational when cattle
production is sufficiently profitable to justify the conversion costs and the opportunity costs of
forestry production. The objective of the land owner is assumed to be profit maximization and thus
the land conversion decision will be made to maximize the joint forestry and cattle benefits.
Following Kamien and Schwartz (1981), USDA (1983), Hardie (1984), Silberberg (1990), and
Parks4 (1995), the land owner will maximize the present net value of the net benefits from forestry
and cattle by choosing a conversion level, c,, to maximize
00
PVNB= f [(F,,pa) Xc, + (C,, p,,a) l dt .
0
Where n, (Ct, P, a) is the rent associated with a hectare of pasture, ,f (Ft, P, a) is the rent associated
with a hectare of forest, and Xc, is the cost associated with converting a hectare of forest to pasture.
The flow of land between forestry and pasture is defined as c,; and constraints to conversion may
occur when inputs to production such as labor and capital are in short supply, leading to conversion
limits: 0
4Although the optimal control theory is well developed in economic literature this section relies heavily
on a similar exposition of this issue by Parks (1995).
|
RSS
HIDE
