It has been argued that a reduction in tropical deforestation rates would be a relatively cheap way of curbing global CO2 emissions compared to the cost of reducing fossil-fuel consumption in the industrialized world (Nordhaus 1991; FACE 1993; Schneider 1993; Kolk 1996). To assess whether this is a reasonable argument one would need better estimates of carbon emissions from deforestation as well as better estimates of the benefits of deforestation.

In chapter 6 we tried to assess the benefits of deforestation in terms of increased rural and urban output. In this chapter we attempt to provide estimates of the rate of carbon emissions arising from land use change in the Brazilian Amazon. We will show that even when we only include the period of most rapid development in the region (1970–1985), our estimates are considerably lower than other estimates found in the literature. This is owing to two factors. First, we take into account the uneven spatial distribution of deforestation: lower clearing costs cause deforestation to take place in the most accessible and least dense forests as long as these are available. Second, we take into account the considerable secondary regrowth and the carbon sequestration that takes place when previously cleared land is abandoned.

The remainder of the chapter is organized as follows. The next section describes the carbon emissions model and the parameter values used for simulations.