THE THIRD HORSEMAN MOUNTS TO RIDE: FAMINE 199|2/93 (Mozambique and Southern Africa ) RSPDOC WWflON SUMMARY INTRODUCTION ^ G ^ BN r - guTDC Irib: 1. The current summer crop year in Mozambique and South/Southern Africa is the worst since 1983; is probably worse than 1983 and may be the worst since early 1920's. "The harvest is gathered / the summer is over and we are not saved" as Jeremiah put it. No rains now would be in time to avert disaster - the .harvests are already burnt out stunted and withered. 2. There can be no precision but the loss of grain equivalent (grain plus root crops plus fruits and vegetables plus plantains) may well be 750,000 to 1,000,000 tones or a third of probable 1989/90 food production (including household self provisioning and local barter). That would raise the nominal national food deficit (assuming unchanged food aid and parallel imports) from about 25% of basic needs to over 40%. 3. That means famine - mass deaths from hunger such as happened over 1982/83 - 1983/84 (and in Tete in 1988 and Manica in 1988 were barely averted). Reports from South Africa/ Zimbabwe/ Malawi; Eastern Zambia, Southern Tanzania/ Botswana/ Lesotho and Swaziland are all along similar lines as to food shortfalls with Zimbabwe estimating up to 6 million ( half its people) in need of food support versus under 1 million last year. SADCC Food Security Unit reports confirm the national data. 4. Time is short. Present stocks/ food aid in the pipelines/ this year's harvests may take urban areas through September and all but the most seriously affected rural through October or November. But . from formal food aid request to approval to shipment to receipt ex port rarely requires less than four months and movement up country ( especially to Manica-Tete-Niassa-Eastern Nampula which this year cannot be supplied from Zimbabwe because of its deficit) takes another month to six weeks. Serious "quantified needs and logistics programmes/requests must get to ' donors by late April if we are to avert drought. 5. We should now be beating the tocsin the loud alarm that the third horseman of the apocalypse (famine) is mounting to ride out - and that Cooperation now is necessary to marshall forces to prevent his killing scores of thousands in Mozambique and up to half a million souls in.Southern Africa. No such loud alarm is yet audible. 6. What is happening is Nireaucrats covering their backs. For example WFP has rather quietly noted the shortfalls. It has 302838331V i" 8 SEP ^ said if not met thousands will die. But it does not go on to demand timely action. It folds its hands saying are distracted / short of food / won't give more and anyhow ^ logistically couldn't move it if it came. This is Pontius Pilate washing his hands of innocent blood and saying "see you to it". It is an obscene exercise in 'Pass the corpses.' We (and other Southern African governments and SADCC on our behalf) must take the initiative. They are our brother , and sisters and parents and children at peril of their lives. They will be our corpses to bury. . We have the duty to rage against what is in grave danger of happening and demonstrate how it can be averted. Food Security in Commerce has much data from at least most provinces. It shows why crops have failed and has numbers for output losses outside family sector. There are rough % losses for family sector. Food security DPCCN/PM's/Finance/Planing need to: a. Convert family sector losses to approximate numbers by Province; b. Compute additional (to present emergency needs) food aid requirements; c. Provide timetable for delivery to avert disaster and; d. Logistical Plan for getting food aid up country (including lorries/spares/fuel and spares for coastal vessels). Some coordination with( especially ) Zimbabwe/ Malawi / Zambia, Tanzania / Swaziland is needed: ' ———-- —- a. a joint SADCC< approach (to complement national ones) would be useful; b. if we need 500/000/750/000 tones; Zimbabwe 750 / 00/1,000 / 000; Swaziland 150,000/200 / 000; Malawi 300/000/500,000 (above refugee food aid); Tanzania 250,000; Zambia 250,000 (unclear as western Zambia harvest may be very good-) then the 2,200.000/3,000.000 tonnes extra grain has to be divided among access ports/corridors . Despite WFP; views Durban is no answer -RSA imports could well fully utilize its slack. But: Dar 500,000 - 750/000 Nacala 500/000 - 750,000 (subject to Nampula/ Intralagos road/rail link) ^ Beira 1/000,000 4k-Maputo 1,000/ 000 '" --'• may be practicable if planned out now 10. The submissions to donors ( and subsequent meetings with them) '' need to be in April/May or they will be too late to provide food in time. 11. War continues to bw Swd food relief (the fourth horseman rides ^^ picket for the third): a. food convoys (let alone individual lorries ) and trains are subject to attack; b. food depots and godowns outside main towns are even more vulnerable; c. We cannot supply food to our civil population in areas controlled by or cut off by Rename. What can be done to save lives? 12. Convoys/distribution points of "peace" may be worth exploring : a. it could - if Rename agrees - save lives; b. it is internationally popular so Rename may feel forced to agree; c. if Rename refuses ( directly or by delaying tactics) they brand themselves as still the same old Bandidos . 13. The main outlines could bei a. No attacks on or looting—of. marked relief food trains-— vessels - lorries - planes - distribution points; b. International observers (sayWFP, UNICEF , UNDP, WHO and - if unavoidable - ICRC) to see that "a" is honoured and that marked vehicles/depots are food and food and drugs only. c. Deliveries to Renamo controlled or cut off areas (via Mozambique ports or via Malawi) for civilians would be allowed. Observers ("b") would see that food was given to women ( who do the cooking) or to women/children at vaccination/basic health clinics. They ( observers ) would report on violations ( Mozambique 's too if so foolish as to line up troops at depot to collect food' ) and if repeated advise halt to deliveries . 14. Yes, some food would be diverted to Rename' s bandidos. Nothing can prevent that^- after all their wives and sisters and mothers are women who wi-11 collect too). Buf' if there is a severe food shortage the bandidos grab all there is and let the civilians starve (as we know from conditions in newly liberated areas). So most of gain will be to civilians and /' there save innocent lives. 15. To be accepted such a proposal needs to be cleared with UNDPWFP-UNICEF-WHO and launched publicly (ie at st. Brigida one day/AIM release and goovernment press conference next). 16. Distribution must be simple. In severely drought stricken rural areas the only administrative.ly practicable approach is to have a rough and ready household list per depot and issue ration to senior woman in each household. Precise tests to find the few who need none or the portion who need less would prevent distributing any. In less severely affected districts could try half allocation to each household. 17. The balance of this memo looks at: i. requirements ii. logistics (including regional ramifications) iii. delivery administration ( including "peace points') iv. approach to donors (including regional coordination ) in more detail. 18. This memo is not self activating. Nor can DNP appropriately be land unit (as opposed 'to key supporter). Food security/DPCCN/Provincial Calamities Commissions are the keys. To fit them into the lock to open food flow door will require high level - high-pr-iority - sustained political initiative. Requirements ^ 19. Any estimated food (im) balance for Mozambique is exceedingly tentative. No meaningful family sector output, population in Mozambique or range/median/mean/mode calorie deficiency estimates of an empirically derived nature exist. At best one has qualitative estimates pieced together from scraps of empirical data/other observations. However, even rough estimates are better than none ( or than precisely wrong projections on implausible premises from dubious bases). a. the probable 1990 population of Mozambique in Mozambique was of order of 12,5 million. b. the grain equivalent for a minimum- adequate diet is 360 kg per person per year. c. the average food djj^icit in Mozambique in 1990 was 25% (true famine levels are the other of 30% - Mozambique was above that point albeit precariously -so). d. food aid (all items) was of the order of 500,000 tonnes grain equivalent and "parallel'/commercial imports'' ! 125,000 tonnes. ' e. ' 1989/90 was, on average, a reasonable to above average crop year so far as weather was concerned. 20. In that case: i. food required 360 kg x 12/500,000 = 4/500,000 t ii. food kg deficit (25%) = 1/125,000 t iii. food imports = 625,000 t Aid (500,000) Commercial/parallel (125/000) v. Domestic production * = 2 / 750,000 t Urban (Zonas Verdes) 150/000 Household consumed ( 50/000) Commercialized (100/000) Rural 2/600,000 Household consumed (2,100,000) ' Commercialized ( 500/000) (Including parallel and local) r ' Urban 'Balance' '~ i. food required ^ Rural 360 kg x 3,000,000 = 1,080,000 t 3,420,000 ii. food Deficit (12%) 130,000 ** 995/000 ( 29%) iii. Imports 400,000 ** 225 / 000 iv. Domestic Production 550,000 2/200/000 Urban (15/000) Rural (400,000) 2,200,000 Excludes perhaps 50,000 t crossborder exports. Largely Tete to Malawi. ** Rural 29% average gap looks high by - say - 3% (100/000 - 125,0001) given absence widespread famine. Probable explanation unrecorded/ Unestimated border transfers Zimbabwe, Malawi, ' Tanzania.  J< 21 The 1992 (1991/92 weather or crop year) drought damage estimates require analysis from provincial calamities commission data. On quick look at Manica data the loss is likely to be: I Low Estimate Urban Household consumed 10,000 (20%) 40/000 Commercialized 30,000 (30%) Rural Household consumed 525/000(25%) 725/000 Commercialized 200,000 (40%) Total Loss 765,000 High Estimate Urban Household consumed 12/500 (25%) 52/500 Commercialized 40/000 ( 40%) Rural Household consumed 630/000 (30%) 930/000 Commercialized 300,000 (60%) 982,500 Note: the detailed data from Manica .are on estimated private-company-coop losses which relate primarily to non-family sector portion of commercialized. They are much lower than total output loss but are higher % loss than' the average (ie family sector some what less drought vulnerable). 22. The disaster in Manica is of a degree that combined w-ith - to date - less detailed reports from Sofala ( and cross border reports Zimbabwe/E.Zambia/Malawi:) it is 'safe' to project 'high estimate' losses for Southern/central 6. provinces. Quelimane and Nampula are possibly less affected; Niassa must be badly hit judging by Southern Tanzania reports; Mueda plateau core of Cabo Delgado may be less drought hit. Further cassava is much less drought prone than grain and these four provinces have higher cassava shave in output so should be less hit. They may be on 'low estimate' loss scenario. 23. This total losses are unlikely to be below 750 ,000 tonnes and the most plausible (least implausible) single guess is - say - 875,000 tonnes (averaging hi and low). Roughly this implies: (Low Estimate) Ar  J«< Increased Urban deficit - 150/000 Lower Production 40,000 Lower Rural Purchases 160/ 000 /• Less (Larger Parallel Impunts) (50,000) Increased rural deficit 515,000 Lower Production 750,000 Less Reduced Sales to (160,000) Urban Less increased cross Border Transfers (50/000) Note The higher imports/transfers are very plausible. But as the whole region is food short no serious size estimate is possible. 24. This 'calculation ' implies Mozambique needs 1992/93 additional food aid deliveries of 665,000 tonnes about 22% to be sold ( or given selectively ) in urban areas and 78% to be given (as crop losses have destroyed the income/entitlements of households needing food) in rural. 25. Allowing for present stocks, domestic purchases, scheduled aid flows the above 665,000 tonnes is needed at roughly 100,000 t. month port arrivals from September 1992 through March 1993 inclusive. / 26. Mote - it will be asserted that 665/000 tonnes additional is impossible because "most Moz-ambican food is food aid now." Note so. About 10% of calories needed are food aid; 70% domestic production; 20% ' missing' (not to be cater). Recorded grain trade may well ' be 350/000 aid/50 to 75/000 local so 80% odd aid but a) not all food is grain; b) not all ( or even most) gr.ain is sold since grain producing family sector eats most of what it produces; c) local sales and 'parallel' domestic and imported ones are not captured in recorded data. 27. Tactically starting with 750,000/1,000,000 tonne shortfall and seeking 875,000 tonnes grain equivalent (say 80% grain, 10% sugar, 8% vegetable oil 2% milk powder) might be best. Donors always try to beat down estimates. ( Our true estimates for Emergency usually topped 1 million t. even on an inappropriate definition of deslocado need and affected numbers so could legitimately have been 1,250,000 t plus. Donors wanted us to cut to 600/000 - 700,000 tonnes which they'd then pledge 'fully' and deliver up to 80%. It is to be feared this odd insensitivity analysis/dicing with recipient death approach will continue so seeking 875,000, which would be 'safe' level for us is least unlikely way to get 500,000 to 550,000 to 550,000 which is bare ag.nimum to avert mass famine.) & . Logistics 28. The needs should de: /- a. broken down by province and district bmergaggye^atiwa . ( at province/district levels) with present 29. From that logistical needs/routs can be articulated: a. arrival by international ports (Maputo, Beira, Nacala) b. coastal reshipping ( using Nacala transhipment for range north of Beira and Maputo for Xai-Xai - Inhambane range) c. rail (including Beira corridor to Chimoio for Manica/Tete which can no longer be supplied from Zimbabwe and Nacala -Nampula-Lichinga line for Niassa which cannot this year be supplied from Malawi) d. road ( from port or rail point) c. depots (more needed to serve sector familial in 15 to 25 km radius around depot to complement existing camp and urban networks ) 30. No new port, rail or main road projects can be done by September. But (except for secondary/tertiary ports) rail and port capacity is growing and above present use levels. Bottleneck likely to be lorries - spares - fuel. Need to estimate how many and propose ways for donors to channel to transport enterprises ( public or private) promptly . A package of quick, bottleneck breaking small road and bridge patchup work ( -if ---data from CPP's can be secured ) ' which can be completed by September/October might also be useful. 31. Most SADCC states will need additional food imports. Given its limited capacity and efficiency Durban will not be available if RSA needs to import 1,000,000 t for its own use. Therefore Mozambique should seek to have SADCC/SATCC coordinate a food logistics plan for region: Country Maputo Beira Nacala Par es Salam Namibia M'z'mb. 300 / 000 200 / 000 100 / 000 Tanzania - - - 250,000 Zambia - - - 250/000 Malawi - - 350 / 000 50,000 Z'b 'we 400,000 350,000 - .-Swazi. 150/000 - - - Botswana 25,000 - - - 75,000 Total 875 / 000 55 .0,000 450,000 550,000 • 75,0000 .*• Note: Additlional grain (or other food) tonnage. Mid point of t 20% to - 20% range. 32. Present data suggest ports could handle that tonnage. The Zimbabwe ( Maputo) cargo might have to go largely via Ressano Garcia depending on September 1992 state of Limpopo Valley line rehabilitation. Beira rail line can handle added cargo as can Nacala line to just past Nampula. Unless line to Entrelagos complete by September upgraded road link Nampula/Malawi needed. Delivery Administration 33. Urban delivery administration relatively straight forward. Put on the order of 150,000 - 175/000 tonnes more grain (and vegetable oil and sugar) via commercio's wholesale/retail networks. In Maputo/Beira seek to increase abastecimento quotas (or fulfillment thereof) whatever long term plans. A physical food scarcity period is the wrong time to dismantle fair price shop chain! 34. Rural delivery administration is harder: a. it needs to be free - crop losses wipe out cash income and entitlement to own grown crops to cat so recipients cannot pay; b. however , if clear decision "ruined" and "damaged" crops along zonal or district lines could have i) full ration distribution free in "ruined" and ii) half ration in damaged; c. a rough household roster is needed for each deport's area with senior woman in each household eligible to collect fortnightly (preferably) or monthly/ presumably on a rota system so not all bucked on 4 or 6 days; d. each household (or each non salario household) eligible. No attempt to sort out 5% to 15% who could pay will prove practical; e. distribution from deports near ( radios 15 km to 25 km maximum) family sector households so they can stay on land but trek in twice a month for rations. Logistical exercise needed on: i. numbers required/where ii. how may can use deslocados camp godowns (storage/record^keeping) but separate distribution point outside c%rap  4> iii. how to set up new depots/stores. 35. Routes and depots of 'peace' would greatly facilitate wholes exercise and make possible saving civilians under Rename control or isolated by them. Proposal at Paras 11-12-13-14-15 pertains. Evidently it requires high level political consideration before any further action possible or elaboration/liaison purposeful. Approach to Donors 36. Clearly action depends ultimately on donor pledges, deliveries. 37. Whether we get them depends hew forcefully, competently and speedily we put our case. 38. Desirable timetable: a. Data collection/analysis March 1-31 b. High level broad outline March 21 c. Technical backup plus request papers to donors April 15 d. Conference with main donors/UN agencies ( USA-EEC-Nordics- May 10 WFP-UNDP-UNICEF ) to negotiate pledges e. Speeding up delivery existing ( Agreed May 10 th pledged food aid or sooner bilaterarly ) —--f. Deliveries begin September 15 g. Initial consultations SADCC (food Security) SATCC on - i. possible joint approach to donors ii. coordinated logistics plan for use Namibia (Botswana only)/ Dar es Salam, Nacala, Beira, Maputo March (food security should initiate) h. Formal logistics regional porta/rail) meeting under SATCC Food security unit SADCC auspices April 15 i. SADCC regional dro^ ht paper incorporation/highlighting national presented by delegation (i/rtcluding each interested member state) at workshops in Brussels and New York . April 20 - May 5. y Envoi 39. If we act beginning now full scale famine can be averted. The size of deficit, the slowness of donor and the capacity limits of our own bureaucracies , the limited left (we could and should have been at this stage by February 1), the real logistical problems and the war mean some lives will be lost. But that is no excuse not to claw.back from the edge of the grave as many human beings as we can. "Do not go silent into that good night. But rage, rage against the dying of the light." 40. We need to consider the interaction of drought/famine relief and reconstruction (especially rural livelihood redevelopment): a. this is not a good time to encourage refugees to return on their own; b. only if UNHCR has food and logistics to support (feed) returnees to March/April 1993 harvest are returns over March/October 1992 at all prudent ( after October too late for clearing - planting - getting March/April 1993 crop); c. if we do set up a more articulated delivery system beyond camps to selected villages this should be used as basis of 1993/1997 Rural Livelihood Rehabilitation Programme delivery system. Probably 1/500 (at most 2/000) depots are needed to allow households to make monthly or fortnightly pickups of food, (and initial tool and input kit pickups )- while staying on/recreating their homesteads. ( That assumes 15 to 25 Km practicabi maximum one way trip depending on terrain.) It would in fact be a major step to breaking the delivery bottleneck as Saude units (and Educacao once it halts reducing and begins raising rural effort), agriculture, small public works, traders could locate near/far out from these depots. Reginald H Green SSPA/DNP/CNP Maputo 14-17/II/1992 ] TheInstitute of Development Studies 1 fnl ^^ B ^ ' |!^ MP iH BJB 2 ^ I attneUniversity of Sussex || JCUBUQiS Brighton BN1 9 RE England Telephone: Brighton (0273) 606261 Telex: 877997 IDSBTN G Fax: (0273) 621202 Direct Line (0273) (,yi(,47 Simon Maxwell 28 April , 1992 Herewith an 'update' on "Tocsin"! Feel free to Xerox to help keep up momentum of NGO pressures which seems to be working - especially re EEC. Not - (a.) On post 1975 cycles likely 1992/3 and 1993/4 will be poor years. (b.) Major seed problem for 1992 plantings especially as normal (RSA-Z'we) sources knocked out. Needed by October at latest, (c.) Major drinking water crisis (deaths to date are from this more than food). For example, 200 plus miles of Limpopo River are just not longer there. Sadly,