starting in March/April.   At this point the Government of Tanzania banned the export of maize.

The real emergency, however, developed later in the year when the main July/August harvest

failed in many areas due to poor rainfall.  FSD's preliminary forecast in May had already estimated

the national cereal deficit for the 1996/97 production year at over half a million tons  (the worst

shortfall since the 1984/5 drought), and in August it released a revised forecast of a staggering

766,000 MT deficit.  Already (in July) a joint UN/GoT assessment mission (with FEWS

participation) had warned of an  imminent food crisis  [ref 11] in nine Regions, and Regional

governments had started to send food aid requests to the PMO.  On September 15 the GoT

declared an emergency, appealing to donors for 76,000 MT of food aid (10% of the forecast deficit)

and waiving the import tax on maize until December.  Donor countries immediately pledged food

aid in support, and a new WFP Emergency Operation (EMOP 5889) was approved by Rome on

November 6.   After a planning period including a village targeting exercise (see below),

distributions began on December 8 but were, ironically, soon disrupted by heavy rains (the onset

of El Nino) which damaged roads, bridges and railways 

10

.  Mainly because of these logistical

problems, the total quantity of food distributed during the EMOP was substantially less than

planned (at around 30,000 MT). Nevertheless, distributions were stopped at the end of April in

response to improving conditions and a good 

vuli  

harvest.  Food aid remaining in hand at the end

of free distributions was to be used for post emergency self help and school feeding projects

[refs 44, 45].

A number of targeting issues arise.  Firstly, the types of systematic and analysable information

available for national (stage 1) targeting were fairly limited.  While it was clear which areas had

suffered from poor rainfall and therefore reduced harvests, very limited information was available

at decision making points  in Dar Es Salaam about the likely 

impacts

 of these losses.  The focus of

the information providing system was very much on quantifying food crop production, especially

cereals.  There was no systematic way to take account of the relative importance of these crops in

the diets and economies of different areas, of people's livelihoods and incomes, and generally their

ability to  cope  with the effects of poor grain harvests.   Decision makers and information 

providers were aware of the importance of these factors, and made efforts to take account  of them

during the planning and implementation of the relief operation.  It is, nevertheless, a critical gap in

the national monitoring system that it is not able to place crop forecasts in a broader food security

context from the beginning.  FSD's Food Security Bulletins do include a frequency count of

reported   coping strategies  in food deficit areas, using a standard list of five activities 

11

 plus a

category of   other strategies  for the whole country.  However, these are too standardised to be

very useful (e.g., pastoralist areas register 100%  engaging in non agricultural activities ), and

cannot be meaningfully interpreted at this level.

The numbers of  affected  people reported and compiled by government channels from the village

through Districts and Regions to the PMO were not a reliable planning basis for targeting scarce

resources, for reasons illustrated by the example of Arusha below.

A nutrition survey was funded by UNICEF  [TFNC, ref 38], but the report was not available in time

to influence targeting decisions.

Joint  Assessment Missions initiated by the government and UN, with varying participation from

other agencies (including FEWS), were important in triggering national level decisions in this

context of large information gaps and uncertainty about the scale of the problem.  They verified

and prioritized problem areas,  put planning figures on the numbers of people and quantities of

10

 The El Nino floods in Tanzania (unlike Kenya) were primarily a short term logistical problem, temporarily

cutting off market access as well as relief distribution.  Some crops were certainly lost and cereal production

was reduced overall,  but other crops (rice, banana, roots and tubers) benefited: thus the food security

impacts were mixed, and no large scale disasters resulted.

11

  Selling cash crops, buying food crops, selling livestock, restricting use of food to food only, and engaging

in non agricultural activities. [ref 22]

A 6