therefore also qualify as technological progress. Based on this definition, intro-
ducing agroforestry practices in most cases would qualify as technological
progress. Technological progress often is linked to agricultural intensification,
defined as more inputs (or output) per hectare, but the concepts are different.
Technological progress may or may not imply intensification, and intensifica-
tion can take place (and often does) without any change in the underlying
technology.
Agroforestry may be introduced at various stages in agricultural develop-
ment. In Table 5.1 we have singled out three typical cases of agroforestry adop-
tion. Because the cases differ, they will help structure the discussion about the
likely impact of agroforestry on deforestation.
The first case is when tree crops are introduced into shifting cultivation
systems, typically fruit or multipurpose trees. They tend to increase the labor
inputs per hectare and are more labor intensive than shifting cultivation or
pastures but not more than continuous cropping of annuals. The main driver
behind the development is population growth (higher land scarcity and the
need to provide more food from a limited land area), and agroforestry is a low-
cost intensification in response to the need to supplement subsistence food
production (e.g., fruits or protein banks as fodder supplement for cattle) and
products traditionally collected from the forests.
The context for the second case is in many respects similar to the first one
in that land is abundant and the forest frontier is still open. The principal dif-
ference is that the driver behind the adoption of agroforestry is the desire to
produce commercial tree crops for an outside market. The trigger can either
be new market outlets (e.g., a new road) or a new technology or production
system being introduced by government extension agencies, commercial com-
panies, or entrepreneurial individuals. Commercial tree crops can be intro-
duced and modify existing systems, such as rubber agroforestry in Indonesia,
which developed from the introduction of rubber trees into the traditional
shifting cultivation system (see Chapter 10, this volume). Eventually, com-
mercial tree crops might become so dominant that the system cannot be
classified as agroforestry anymore. Sunderlin et al. (2001) found evidence of
such dominance in the case of rubber, cocoa, and coffee in Indonesia. An
extensive survey of more than 1,000 households in the outer islands of
Indonesia suggested that among those clearing forests, almost one-third did so
for sedentary agriculture of mainly tree crops (more than half chose rubber),
another third for 1–2 years of annual food crops only, and the remaining com-
bined the two.
The third case presents a different situation in which scarcities of land and
forest products are major driving factors for implementing agroforestry on
farmland to provide forest products. The demand for these forest products
typically is from local or regional markets, not international ones (unlike the
second case, in which markets can be national or international). In this
- Is Agroforestry Likely to Reduce Deforestation? 91