It is argued in this paper that two fundamental economic processes pre
vent resource-poor farmers in tropical countries from managing soil ca
rbon in a sustainable manner. The first process is related to the fact
that soil carbon and tropical forests are part of the natural capital
of these countries and of the world community. As a consequence, the
interests of resource-poor farmers in tropical countries, of these cou
ntries themselves and of the world community conflict. This implies th
at levels of adoption of sustainable soil carbon management practices
which are optimal from the perspective of resource-poor farmers are su
b-optimal from a regional and global perspective. The second process r
egards the nature of sustainable soil management practices. These prac
tices are investments in natural capital which bring about net benefit
s to farmers only after four to six years. Absolute poverty levels in
tropical countries make it very difficult for farmers to undertake suc
h investments. It follows that even perfectly informed and rational re
source-poor farmers will not voluntarily adopt socially optimal levels
of soil carbon management in tropical countries. Policy interventions
are a means of ensuring that soil carbon is managed in such a sociall
y optimal and sustainable fashion in these countries. Two principles a
re proposed for developing effective, equitable and appropriate policy
options. The first is the beneficiary-compensates principle, which re
quires that society in tropical countries and in industrialized countr
ies should compensate resource-poor farmers in tropical countries for
adopting soil carbon management practices. The second principle is tha
t international and national policy options need to be well articulate
d and that sets of complementary policies should be put in place for s
eater effectiveness. Policies alleviating rural poverty and pressures
to deforest are as necessary as policies specifically targeted at soil
carbon management. Finally, research priorities for soil and biologic
al scientists so derived from the analysis. These priorities necessita
te the creation of interdisciplinary teams of soil, biological and soc
ial scientists. This is perhaps an even greater challenge for the scie
ntific community than the achievement of the research agenda itself. (
C) 1997 Elsevier Science B.V.