Cost of feeding refugees
cost of feeding a refugee? Michael H.N. Golden 09.07.98
Cost of feeding refugees André Briend 11.07.98
cost of feeding refugees Michael H.N. Golden 11.07.98
Cost to feed refugees Serge DEPOTTER 23.07.98
energy density of relief foods Michael H.N. Golden 14.08.97
Food Aid Pieter Dijkhuizen 08.09.97

Subject: cost of feeding a refugee?

Date: Wed, 9 Jul 1997 12:29:40 +0100 (BST)

From: "Michael H.N. Golden" <>


I have heard in meetings that the expendature on the general ration for a refugee in Africa is about USD 0.20 per head per day. And that the general ration supplied in Former Yugoslavia was about USD 5.0 per head per day.

Does anyone know where we can obtain data on the estimated cost of feeding a refugee with the general ration in different countries/regions - and the breakdown into costs for food itself /transport/ Admin etc?

Are there any similar figures (in the same format) for the cost per-head-per-day of supplementary feeding programs and of therapeutic-feeding programs?.

Where there is an inadequate general ration then a proportion of recipients will deteriorate to require supplementary feeding and similarly for the supplementary feeding failures going on to death or thereapeutic feeding. I suspect that the cost of improvement of the general ration would be less than saving on the general ration and mounting a more expensive supplemental program for those that fall through the net and similarly for better supplemental feeding to reduce the requirement for Therapeutic feeding. If true, it would be useful to know the quantum of increase in the general ration/supplemental feeding cost that could be justified on economic ground alone (there are other very cogent grounds) assuming that there was a resultant percentage reduction in the deterioration to the next stage. This would then guide the sort of stratagies that we can develop to improve the general ration/ supplemental feeding.

Is anyone working upon mathematical models of population changes in nutritional state and the effect of intervention at various stages and in various ways?

Where is the literature on these aspects of refugee nutritional management?

Best wishes,


Prof. Michael H.N.Golden

Fri, 11 Jul 1997 10:15:42 +0200 (METDST)

From: (Andre' BRIEND)

Subject: cost of feeding refugees


Dear Mike,

I am afraid you don't have much feedback to your queries re: cost of feeding refugees. I suspect nobody has the data needed to make these estimations.

I had for a long time this question in my mind, and already did some calculations on the back of an enveloppe. Some conclusions:

1- You save money if you stint on general ration, even if this is at the cost of opening TFC's.

Usually, in non critical situations, you rarely have more than 1% of children in tfc. This makes 0.2% of the population of a camp in a tfc, assuming children below 5 are +/- 20% of the population. If opening a tfc is the price to pay if you want to save 10 % of cost of general ration, you can spend a lot on tfc, and still save money. No suprise if this option is often chosen. You save money by improving the general ration only when number of children in TFC becomes huge. Note that non food costs (including suffering) involved in opening tfc are usually not supported by the agency providing the general ration. TFC's on TV are 'sexy', and there is no problem to find separate donors for them.

2- You save lives if you have a good general ration, even without TFCs.

As you know, mortality increases even for moderate weight deficits. Moreover, Type I deficient children will have increased mortality (see vitamin A story, see iodine, infant beri beri ... and I am sure more to come) even when not included in the 'undernourished' by anthropometry and tagged to receive extra food. Hence targeted feeding programmes based on selecting children with anthropometry (either sfc or tfc) will be less effective to prevent deaths than a general improvement of the situation.

I leave it to you to guess which option (saving money or saving lives) is the most frequently chosen by the international donor community.



Dr. André Briend

Subject: cost of feeding refugees

Date: Fri, 11 Jul 1997 12:55:12 +0100 (BST)

From: "Michael H.N. Golden" <>


Dear André

Thank you for your response. You are correct - the RNIS and WFP data-bases that I have access to do not give the information needed. There seems to be no data on the cost of supplemental or therapeutic feeding.

The concept of type I and type II nutrients is important in helping us understand why there is a poor relationship between mortality and anthropometry (which has hitherto been ascribed solely to infectious disease), and the problems with targeting based upon some index of wasting only. But these arguements come after we have the basic information for a simple model.

>From my "back of the envelope" calculations I concluded that the relative costs were critically dependent upon the actual magnitude of the different components of a relief operation.

Thus, let us assume:

cost of general ration 1 unit per head per day

dry supplemental ration 4 units /h/d

therapeutic feeding 20 units /h/d

population 20% children 6-59 months.

prevalence of wasting

moderate 20%

severe 2%

Coverage 100%


Then if only general ration is given the cost will be 1*population, whereas if there are also a supplemental and therapeutic feeding programs the cost will be 1.2*population or a 20% increase in the total costs. Which is far from as trivial as your calculations suggest.

Nevertheless, it begins to give me a ball-park in which to work. If these figures are approximately correct - the question becomes "if by increasing the cost of the general ration by about 20% through adding vitamins and minerals (holistically) and diversification - a) could we within this cost achieve the RDA for all nutrients? b) what would be the reduction in the rate of progression to moderate and severe malnutrion? c) what would be the additional benefit of prevention of mortality in the mild/moderate group becasue we were supplying adequate type I nutrients (deficiency not directly related to anthropometric criteria) etc? d) what would be the additional benefit to the 80% of the population not in the 5-59 months age range that we now ignore? e) to what extent would having a nutritionally healthy population help towards progression to self-reliance and shorten the period for which the people need to be fed? And several more critical but more difficult considerations to quantify.

In a complex model of the effects of refugee feeding stratagies (on population dynamics-health-productivity) the effects of each of these factors should be included, instead of the simplistic calculations that we have done. When we have a mathematical model the research community needs to set about obtaining extimates of the magnitude of the variables and exponents of the parameters - in a similar way as Roy Anderson's group has done for infectious disease modelling, and departments of agricultural economics model the effect of changes in farm prices.

Of course "coverage" is never 100% - it is closer to 50%, so the actual increase in costs fo refugee feeding would be about 10% (but it SHOULD be 100% and the increase in cost should be 20% not 10%). - this begs the question of what happens to the 50% of children that are malnourished and do not appear for supplemental feeding/therapeutic feeding (presumably many of them die) - finding them would increase the costs of the programs enormously (cost of finding and of treating when they are found) - So I now ask - by improvement of the general ration would the 50% of children that are malnourished and never appear at a SFC or TFC be "covered".

I suspect that improvement of the general ration is not only a more cost effective stratagy for all sections of the population, but also for the 5-59 month old children (at least for the 50% that the special programs do not reach!)

Would any of the economists on the list care to comment? (Roy, Bea etc)

Best wishes,


Prof. Michael H.N.Golden

Date: Wed, 23 Jul 1997 10:24:23 +0100

From: (Serge DEPOTTER)

Subject: Cost to feed refugees


Hello Mike,

1. Thanks for yr answer

2. Yes I think it would be nice to be included in yr E-mail list so Saskia (MSF in Amsterdam) doesn't have to filter for me

3. I send a msg cc "Costs to feed a refugee" to WFP in Rome on which they send me next answer. I am afraid it doesn't bring us much further as all info is commonly known by everyone....

The prices below (from WFP) show enormous differences in transport depending on the country and the period.

(I remember that during the Goma crisis the transport costs rose with more than 500% because of the high demand from all the NGOs; transport to southern Ethiopia costs 5 times as much than to the northern part because of security problems.)

This means that it may be worth to make a study on this issue. In any case the price of :

- relativ well being

- long term effect of chronicle (sub)malnutrition (WFP rations of 1500-1700 KCal/p/d without pulses are not rare in refugee camps!! last example: Dadaab camps in Kenya) is hard to calculate. International donors often want hard figures and don't consider subjectiv arguments such as "dignity of refugees".

Here follows the answer from WFP

Greetings from Brussels,




>Dear Serge,

From Delphine I got your request for information about this issue. Indeed in WFP we analyse constantly the cost of our operations to find the most economical way to spend our limited resources.

Unfortunately I lack at the moment the time to go in much detail.

The cost of a feeding regime can be calculated as follows:

1. FOB world market cost of commodities of basic food ration (2115 kcal), for example:

400 g maizeflour : USD 240/MT

60 g pulses : USD450/MT

25 g vitA fortified veg.oil : USD 900/MT

60 g fortified blended food : USD320/MT

20 g sugar : USD410/MT

5 g iodized salt : 210/MT

(for locally purchased commodities local prices will apply; ussually these are considerable higher than world market prices)

2. International transport costs, depending on location, ranging from USD50 till several hundred dollars/MT (if applicable)

3. Local transport, handling and distribution costs, also widely ranging up to several hundred dollars/MT, depending on local situation. For air transport the costs can increase till over USD 1000/MT

4. Administrative costs vary between 6 and 13%, depending on the type of operation

5. Losses are small, on average below 1%

On average WFP spends about USD 0.25/day to feed a refugee, however there is quite a cost range, depending on the local situation.

For supplementary feeding a similar calculation applies, however distribution costs in wet feeding are much higher in such case. Hope this is useful, best regards

Pieter Dykhuizen

From: Prof. Michael H.N.Golden

Thu, 14 Aug 1997 17:46:14 +0100 (BST)

Subject: energy density of relief foods


Fellow ngonutters,

Now that we have got a logistician on the list (Bruno Corbe- MSFB), I think we could usefully discuss the way in which we express the cost of food items and how nutrition planners need to interact with logisticians.

When I was writing a report for WHO/UNHCR at the start of the relief effort for Sarajevo, on the ration that should be delivered, I tried to consult the logisticians. The response was "It is your job to decide what has to be sent and our job to get it there" - a clear demarkation with a rejection of dialogue. At that time only 10 ton trucks could get over the mountain passes.

The difference in transporting fat at 9 kcal/g, wheat at 3.2 kcal/g or potatoes at 0.7 kcal/g (the local population wanted potatoes), would clearly be enormous - and there were choices to be made in the "energy density" of the recommended ration. One truck of oil would carry as much energy as nearly 13 trucks of potatoes - but other aspects of storage, pallet size, ease of handling etc were not part of the consideration, although an adequate nutrient content of the diet was of course the first priority.

Without logistic consultation I simply made the diet as energy dense as possible because I perceived that transport was the major constraint in the relief effort. Despite the problems that we had with ascorbate in the diet, the offer of a gift of thousands of tons of apples (0.3 kcal/g) was turned down because of percieved transport/storage consciderations.

I think that this story shows the critical need for nutritionists and logisticians to work very closely together when designing diets for relief efforts; I would like to know the experience of others in this respect.


The units that we cost/purchase/transport/deliver food aid are simple and easily understood by all - usually metric tons - and the food is given out it kilograms. However, we state the requirements in terms of kilocals, not grams of food - and there is a move to express other nutrients in terms of their density per unit energy. I believe that a change in the way that we express the cost of food would have a major effect on our perception of the diets that should be delivered. When the transport costs are added the differences are even greater - it is only by making the calculations do these point become clear.

Food Energy
food cost
maize 240 3.50 69 29
pulse 450 3.00 150 33
oil 900 9.00 100 11
sugar 410 3.95 103 25
blended food 320 4.00 80 25
  food cost
food cost
maize 97 126
pulse 183 217
oil 111 122
sugar 128 153
blended food 105 130

notes: One gigacalorie is equal to 1,000,000 kilocals (the same number of grams in a metric ton). I have not managed to factor in the weight/bulk of packaging material for the different items - clearly important in terms of oil (although the containers have spin-off uses).

- if we were feeding 2100 kcal per person per day one gigacal would feed 476 people. Whereas one metric ton would feed different amounts of people depending upon the nutrient density of that metric ton.

When we express the data per gigacal the impression we get changes quite dramatically - oil which per MT is twice the cost of pulses, is now two thirds the cost of pulses and about the same as sugar.

The cost of transport is based on weight. Thus, the more energy dense the food the cheaper it becomes to transport 2100 kcal to one refugee. The cost of transport varies dramatically.

The minimum is about US$ 50/MT for international transport and US$ 50/ MT for local transport so the calculation I've done with US$100 as the transport cost is about the minimum to be expected. The cost for both these can go up to over US$ 400/MT in more difficult circumstances and the cost of air delivery to over US$1000/MT.

I show, in the table, the cost (food plus transport) of sending one gigacalorie of each of the common relief commodities at either a total transport cost of US$100/MT or US$200/MT : as the cost of transport increases the relative costs of supplying different foods changes so that at around US$160/Mt transport it is cheaper to give people 2100 kcal of oil than it is to give them 2100 kcal of maize. Adding oil into food actually makes it cheaper to feed a refugee. Also the difference in cost of blended food and maize almost dissapears because of the slightly higher energy density of the blended food.

We often have the choice of sending a diet with 15% or 40% of the energy as fat without the level of the other nutrients being diffrent.

The high fat diet SEEMS to be much more expensive when expressed per metric ton, but this is the wrong way to express the costs - it should be in terms of delivering 2100 kcal to an individual - and this depends dramatically on the number of kal in one gram of the food.

Not only should we think of our costs but also the "costs" of carrying energy dense or dilute food by the refugee from the distribution point to their home.

If we expressed the cost per gigacal then the difference in cost of maize (3.5 kcal/g) and a high fat balanced blended food (4.7kcal/g) is much less than when the cost is expressed per metric ton. The examination of costs (per metric ton) by funders might be quite different, with less "sucking of teeth", if we expressed all costs on the basis of delivering one gigacalorie (or even 2100 kilocalories!)

Most diets in the developing world, unlike our diets in the west, are very low in fat - if only partial rations are given and the refugees have to gather/grow some/most of their diet then what they get themselves will be almost devoid of fat and a high fat supplement would be particularly beneficial.

I understand that the cost of oil dramatically depends upon the packaging - bulk oil is much cheaper than oil in 40 gallon drums which is cheaper than 20 or 5 litre tins. These factors can be taken into conscideration in any particular operation when we, as nutritionists, are making choices concerning the diet that should be given - but we must have dialogue and work closely with logisticians and economists when we are designing what diets we should give. Logisticians should realise that we have choices to make that can change the ease of their task enormously.

Sometimes, I think that we should develop some very high energy dense foods where transport is a particular problem - for example where air freight is required. or transport is just difficult at any cost - in South Sudan at Barh El Ghazal there has been almost no program of relief because of the transport problems - no roads, everything has to be hand carried for at least 15 km to reach Barh El Ghazal. It would be madness not to send the most energy dense food ration that we can devise provided that it was nutritionally balanced and adequate. The rations developed to carry across the antartic are not necessarily inappropriate to carry into inaccessible parts of Sudan!

Best wishes,


Prof. Michael H.N.Golden

Mon, 8 Sep 1997 14:16:10 +0200

From: "Pieter Dijkhuizen"<>

Subject: Food Aid


The NGONUT bulletin is an excellent idea; it brings useful up-to-date information, supplied by experts, instantly to end-users. And sometimes it even provides amusement.

In WFP we couldn't restrain a smile recently, when a concerned academician proudly presented the wheel invented again in a contribution on 'energy density of relief foods',

As a matter of fact the World Food Programme uses already since the early 90's energy density as one of its parameters: in the WFP web side food costs are expressed in US$ per MT and kcal.

The selection of an appropriate food basket is actually quite complicated. WFP has a standard procedure, with a multi-disciplinary team of experts in nutrition, gender, logistics, economics, procurement and others, to determine the most appropriate and cost-effective ration and approach. This cannot be explained in short context: WFP has manuals of several hundreds of pages for this.

Many factors are considered in the selection of food baskets, like nutritional value, cultural preferences and taboos, availability of commodities, vulnerability in transport, funding directives, distribution requirements, suitability for fortification and gender appropriateness. The relative importance of the different factors varies with local conditions: energy density is only in a few cases important.

Therefore the suggestion to express food in gigacalories is not practical. Food is measured in weight when procured, transported and distributed. A different connotation by nutritionists will only be confusing. Furthermore oil is only in theory more attractive than in practice: it may contain three times more energy than cereals, but it costs usually five times more, is difficult to transport with high losses, is perishable and has a problematic distribution.

Therefore oil is generally less cost-effective than cereals in food aid rations.

WFP's expertise is based on 30 years experience in delivering over 50 million tons of food aid as the largest food aid agency, responsible for all food aid activities in the UN system, including refugee feeding operations. It has come across almost every conceivable situation in food assistance. WFP has many formal and informal agreements with NGO's. For food aid related issues WFP will on request provide expert advisory and implementing services to serious organisations, whenever possible free, otherwise a low overhead will be charged.

WFP is continuously exploring new avenues to improve quality and cost-effectiveness; at present it is engaged in pilot studies of milling and fortification of cereals at distribution sites.

WFP welcomes proposals with innovative alternatives, but don't waste our time with 'privileged glimpses of the bloody obvious'.


Pieter Dykhuizen

Senior Programme Advisor on Nutrition

World Food Programme (DYKHUIZENatWFP.ORG)