Notwithstanding the importance of agricultural productivity growth through shifts in crop
composition, there is less evidence that food marketing reform has promoted intensification of the
key food crops that constitute the bulk of area cultivated in Africa.11 A major challenge,
therefore, is to design input and output marketing systems that support sustainable increases in
farm productivity growth for the millions of low-input semi-subsistence rural households that
can’t move and have limited capacity to change their crop mix. Policy reform needs to expand its
focus from liberalizing markets to solving the broader problem of how to induce technical
innovation and productivity growth to support structural transformation (Staatz and Ba 1996;
Jayne and Jones 1997; Reardon et al. 1996; Reardon et al. 1997).
So far, agricultural marketing reforms have replaced often unreliable, high-cost, and centralized
forms of state marketing with more open markets that may be competitive but often lacking in
information, infrastructure, and are poorly integrated with other key activities. On the input side,
financial market failures restrict farmers’ access to credit and thus constrain the demand for
productivity-enhancing inputs, which in turn limits private investment in input production and
delivery systems (especially serving remote areas). Poor information available to farmers and
many traders about fertilizer types, qualities and application rates, weak coordination between
importers, wholesalers, and retailers, and levels of scale/scope that are insufficient to reduce unit
costs all depress the use of productivity-enhancing modern inputs. These problems have led to a
contraction of fertilizer use in smallholder areas, for example, in Burkina Faso (Dembélé and
Savadogo 1996) and in Zimbabwe and Zambia (Rusike et al. 1997). Moreover, because of the
elimination of state credit services to smallholders and because of the continuing weaknesses of
informal credit markets in semi-arid Africa (Christensen 1989), farmers who have the cash on
hand for fertilizer and seed are either: (1) those with nonfarm sources of cash income, which
relieves the credit constraint (e.g., see Reardon, Crawford, and Kelly 1994 for African evidence;
Kelly et al. 1996 for Senegal; Savadogo, Reardon, and Pietola 1995 for Burkina Faso; and Clay et
al. 1995 for Rwanda); or (2) in a cash-crop scheme where the farmer generates a sure source of
cash income and can often acquire inputs on credit from the cash crop program to use partially
on food crops (such as in the case of coffee in Rwanda, see Clay et al. 1995; and cotton in Mali,
see Dioné 1989).
The foregoing assessment generally indicates that the transition from controlled production and
distribution systems to more market-oriented economies has had widely variable effects on
investment incentives and economic performance. While policy reform has created modest
successes in some cases, the newly-emerging food systems in most African countries have failed
to produce the anticipated stimulus to agricultural growth and rural welfare. This conclusion is
consistent with the assessment of Latin America by de Janvry, Key, and Sadoulet (1997). As
Reardon et al. (1995) noted, the notion of “getting prices right” has been insufficient to
dramatically raise farm productivity, and thus "the debate should be reopened on identifying cost-
11 A notable exception is rice in Mali's Office du Niger zone, where liberalization of output marketing
and processing, combined with decentralization of management of irrigation perimeters, have led to
substantial intensification and production increases (Cisse 1997).
14