need to be examined in future research are the flow, timing and volume of these
transactions. This research should include questions related to the flow of food from the
rural to urban area, the movement of food among regions, ADDs, EPAs, and local
markets; cross-border flow of food; and the shift in food consumed or marketed over the
calendar year.

ADMARC historically has been the dominant player in the food (primarily maize) market.
To decrease this dominance, broadening and deepening the food market by increasing the
number of traders competing in the market is critical. Widening the maize price band to
reflect import and export parity price will allow greater spread of marketing margins to
attract more traders into the maize market (Gray 1995). Expanding the number of traders
will lower transaction costs and provide better prices to both producers and consumers.
Once the maize market becomes more efficient, with larger numbers of buyers and seller,
a price band may no longer be necessary.

Other means of broadening the food market include the following: 1) Expanding
availability of storage facilities in the countryside possibly through renting or selling of
some of ADMARC's warehouses. On-farm storage should also be expanded. A recent
study indicates that losses in on-farm storage are low, at five to seven percent. 2)
Maintaining and expanding removal of artificial restrictions on trading. A great deal has
already been done in this area but vigilance is needed to maintain these advances. The
only remaining significant trading restriction is on export of maize. This restriction should
be examined closely for removal as soon as possible. Liberalization of food trade on a
regional basis could provide higher producer prices, greater price stability, lower
marketing margins and greater variety of food on the market to benefit consumers. 3)
Markets run on expectations of future events. The action of government plays a
significant role in sending to the market signals that create these expectations. An
announcement of a maize price band, for example, sends signals that establish one type
of expectation in the market. Subsequent government actions that do not support this
announcement, send different and possibly conflicting signals to the market. Conflicting
signals create uncertainty in the market. Uncertainty retards market expansion and
development. Government policy announcements need to be realistic in terms of resources
required and available for implementing the policies announced.

Appropriate food processing at both intermediate and final stages is needed, particularly
in the rural areas. Evidence from neighboring countries (Jayne and Rubey 1993), for
example, indicates that widespread use of small scale hammer-mills can make available a
differentiated variety of food products tailored to low income consumers. In Malawi,
consumers indicated a cost-related preference among differentiated food in a study on ufa
woyera and mgaiwa (Mkandawire 1993). In Mozambique, differentiation of food products
resulted in a greater number of varieties of affordable food on the market thorough use of
local processing and introduction of yellow maize. This variety had a direct and positive
impact on the food security of the poor (Jayne et al., 1995).