Future throughput of assembly markets

Contents - Previous - Next

The future throughput of produce marketed at an assembly market is dependent on how the local production base is likely to grow and what proportion of this increased growth will pass through the market. This will be influenced by a number of factors and it is essential that the nature of existing marketing channels is understood before interventions are made. The precise role of existing produce markets should be established, i.e. the percentage of produce marketed through direct purchase at field level, what goes to rural assembly markets and what is delivered directly to the urban wholesale and retail markets. The calculation should also ideally take account of the losses that will occur in the marketing chain. An illustration of the calculation that will need to be undertaken to obtain the approximate throughput of a local assembly market is shown in Box 3.

Calculation of the present throughput of a rural assembly market

Estimated local production of fruits and vegetables ( in tons)
On-farm use (in tons),
i.e. consumption of produce by rural households
(per caput consumption in kg. x farm population 1000)
+ use for fodder + use for seed + on-farm losses
Net marketable surplus (in tons),
i.e. Gross marketable surplus less estimated marketing losses
Farm gate sales and direct delivery to urban markets (in tons)
Net amount passing through all local assembly markets (in tons)
Percentage going through the specified rural assembly market
Estimated throughput of the specified rural assembly market (tons)
Present and future demand at retail markets

The key indicators which explain the present throughput of a retail market are the market's present share of the retail trade and the consumption of produce. As explained above, the latter is normally measured on a per caput basis. Thus, as a simple illustration, the present annual throughput of a market in tons would be as shown in Box 4. This method should be cross-checked against any estimates of throughput obtained during surveys (i.e. counting the number of stalls and multiplying by an average turnover per stall, expanded by the number of working days per year to obtain an annual total).

To use the information shown in Box 4 to project the future demand for the foodstuffs that will be sold through a retail market will depend on four additional factors:

Calculation of the present throughput of a retail market

Per caput consumption of fruits and vegetables ( in kilos)
Urban population
individual market's percentage share of the overall retail trade
divided by
1000 (to convert to tons)

More detail on how these factors can be used to project demand is given in Chapter 12 of the Wholesale Markets Planning and Design Manual (FAO, 1991). Of these factors, reasonably reliable estimates are usually available only for population growth. However, Census data can often be rather out-of-date and this may be a significant factor when urban growth rates are very rapid.

Adequate macro-economic information on which to make sound estimates of disposable income growth is rarely available. Nevertheless, if indicative income elasticities can be calculated these should be used, as even relatively small increases in urban incomes are likely to produce quite substantial increases in demand and hence in potential market throughput.

Central markets may still dominate retail business in a town or city, even when the population served by them is limited. In Zanzibar, for example, although it might be expected that the catchment area of the central market would be limited to only 20 percent of the total urban population (i.e. the inner urban population), the estimated proportion of produce retailed through the central market is over 70 per cent of the total. However, in Zanzibar's case (which is typical of many urban areas) it would seem probable that as the town continues to grow the suburban markets and other retail outlets will claim a greater share. One way of assessing such changes would be to check the number of planning and building applications that have been received by a local authority for the construction of new suburban supermarkets and other retail facilities.

Given the usual paucity of data and the complex interplay of factors, methods for forecasting future demand should be as simple and robust as possible. An over-sophisticated approach, trying to obtain a firm prediction of throughput, is almost bound to fail. The most practical technique is usually to employ a high and low-case scenario based on the best information that is available. For example, a low-case scenario could assume that demand will grow only at the same pace as population and a high case assume that incomes will also grow. Any information available on urban income elasticities of demand can also be incorporated if they seem to produce realistic results. On that basis, the urban consumption can be projected. The method is illustrated in Box 5.

Calculation of the future annual throughput of a retail market

Projected per caput consumption of fruits and vegetables (in kilograms) (adjusted to account for income change and/or elasticity if data is available) (5 year projection) Assume per caput Consumption of 180 kg. Per caput
divided by divided by
1,000 (to convert to tons) 1,000 (to convert of tons)
times times
Urban population x growth (growth rate as a Decimal number plus 1.0) Population of say 10,00 x 4% growth rate i.e. 1.04 x 1.04 x 1.04 x 1.04 x 1.04
= =
Total urban consumption (tons) Consumption of 2,190 tons
times Times
Individual market's projected future % share of the overall retail trade Share of say 60% (i.e. x 0.6)
= =
Future throughput Future throughput of 1,314 tons

A typical example of a projection is shown in Table 3.7. In the case of the high estimate, consumption of fruit, vegetables and fish will have doubled by 2005 and consumption of meat and chicken will have gone up by about three and a half times. However, retail throughput at the market is only likely to go up in proportion if it retains its overall market share which, in this case. is not probable.

TABLE 3. 7 Zanzibar Town projected daily consumption

Year Urban consumption (tons per day):
Fruit & Veg. Meat Chicken Fish
Low Estimate - Assuming Population Growth of 3.6% Per Annum
1993 85.91 1.16 0.49 11.25
1995 92.21 1.25 0.53 12.07
2000 110.04 1.49 0.63 14.41
2005 131.33 1.77 0.75 17.20
High Estimate - Population Growth of 3.6% Per Annum + Income Growth of 2% Per Annum
1993 85.91 1.16 0.49 11.25
1995 99.14 1.51 0.63 12.70
2000 140.50 2.64 1.07 17.16
2005 196.90 4.26 1.69 23.13

Source: UNCDF Project No. UFT/93/C06 (renovation of Zanzibar Stone Town Market, 1994)

TABLE 3.8 Maseru market: preliminary throughput projections

Design Year Urban population Wholesale market share # Consumption per caput (kg.) Market throughput(tons):
annual av. daily
gross net *
1992 164,000 0.54 125 67.5 11,070 30
2000 300,000 0.54 145 78.3 23,500 64
2010 500,000 0.54 165 89.1 44,500 122

Source: FAO Project No. LES/88/0 10

Notes: # 0.75 (overall share) x 0.9 (household gardens) x 0.8 (supermarkets, etc.)
* gross consumption x wholesale market share reduction factor of 0.54

An alternative method for projecting turnover at a market is simply to use the consumption data and develop realistic targets for increases in consumption based on the changes in the pattern of consumption in neighboring countries. Table 3.8 provides such a projection using consumption data derived from a sample household consumption survey and adopting a single scenario for population growth. The projection also provides a direct calculation of throughput by assuming that the market will attract 75 per cent of the town's overall trade, and by excluding home garden production and a proportion of the produce supplied by importers directly to supermarkets, institutions and hotels. To achieve such a target it may be necessary to enforce market regulations in the urban area and to limit the operation of "informal" markets.

Turnover projections and floor space requirements

As mentioned above, projections of future consumption for an urban area served by a market (see Table 3.7) should ideally be based on two hypotheses: a low and a high estimate. These estimates reflect population and income growth, income elasticity of demand and competition from other marketing channels.

On the basis of these projections the next question that should be addressed is whether the present market can cope with any increased trade in terms of its building floor space (sales area and other uses) and site requirements (assuming that the market will continue to play the same role in the future and will maintain its overall share of retail trade as at present). Equally, for a new market, the projection of floor space requirements will be a key factor in the preparation of any market master plan.

Typically, small-scale retail markets should achieve a sales area turnover in the range of 5 to 20 tons per m/p.a. To accommodate this sales space the overall site needs to have a turnover equivalent of I to 4 tons per m/p.a. This means that the building area should ideally be around 20 per cent of the total site area and, except in exceptional circumstances, (e.g. street markets where the market forms part of the common street system) should not exceed 40 per cent of the site area.

These standards are based on empirical data on the relationship between space requirements and the gross annual turnover. The standards include space for sales and general internal circulation as well as a contingency allowance for daily (and to some extent seasonal) fluctuations in space requirements. The figures also assume an average mix of produce, with fruit and vegetable sales predominating. If a market specialises in a particular commodity, some adjustment to these figures may be necessary. An appropriate range of values that can be used for projections is shown in Box 6.

Sales areas: typical throughput values for retail markets

Type of market Annual throughput
Rural fruit and vegetable open market average 5 tons per m/p.a.
Rural fruit and vegetable all-year assembly market 15- 20 tons per m/p.a.
Seasonal assembly market (assuming 3 month peak) 20- 25 tons per m/p.a.
Urban fruit and vegetable open market 5-10 tons per m/p.a.
Urban fruit and vegetable covered market 15- 20 tons per m/p.a.
Combined urban market (fruit, vegetables, fish, meat) 10- 15 tons per m/p.a.
Specialised fruit market (with fruit in packed cases) 20- 25 tons per m/p.a.
Specialised covered fish, poultry, egg or meat market 5-15 tons per m/p.a.
Specialised retail grain, potato or onion market 10-15 tons per m/p.a.

These standards can be used both for the development of a new market and the assessment of an existing market. If an existing market broadly matches these standards then the sales space should be examined to see if there is scope for it to be used in a more efficient manner in order to meet future demand. No additional space may be required but it would be essential for management practices to be improved in order to meet these targets.

In any case, with the introduction of physical and management improvements the utilization of a market will become more efficient over time, which means that the unit space required for the various activities will therefore gradually decrees - this might be encouraged by changes in rental levels. Thus, a standard of say five tons per m might be used for the initial period of operation for an open market, increasing to ten tons per m with improved management. Table 3.9 shows a projection of floorspace and overall site area using such a technique.

Table 3.9 Maseru market: sales space projections (m)

Design Year Annual throughput (tons) Sales area: Site area:
Space standard (mt/m) Total area (m) Space standard (mt/m) Total area (m)
1992 (initial) 11,070 10 1,107 2.0 5,535
2000 (interim) 23,500 12 1,960 2.5 9,400
2010 (ultimate) 44,500 15 2,970 3.0 15,170

Source: FAO Project NO. LES/88/010

Although the standards in Box 6 represent reasonable average values, they need to be used with some caution. Some markets have very strongly peaked trading patterns either on a weekly or seasonal basis. If this is the case, the projected space requirements may only be sufficient for average conditions. Peak daily fluctuations (e.g. between a normal weekday and the weekend) and peak seasonal variations (with, for example an assembly market or an urban retail market that also has some wholesale trade) may necessitate the use of additional space. It may be uneconomic to provide this to the same standard as the rest of the market facilities and some special provision may need to be made, such as the use of a nearby car park or an overspill area, located outside the immediate market area, as an open sales area.

Determining ovrall site requirements and ancillary facilities

The method for estimating the overall space requirements for a market site is shown in Box 7. In addition to the sales space an allowance needs to be made for other ancillary buildings that may need to be accommodated on the market site. For a normal small-scale urban or rural retail market this is likely to consist of a market manager's office, toilets and security/cleaning arrangements. Suitable space standards for such facilities are also shown in Box 7

Specialised facilities may also be required. These facilities might include:

Estimating overall space requirements for a basic retail market

Projected annual market throughput (see Box 5)
divided by
Sales space standard tons per m/p.a. (see Box 6)
Total sales space requirements (covered and open)
Office space @ 10 to 15 m per office worker
Cleaner's store plus guardhouse @ say 10 m
Toilets @ 2 m per 25 market employees (male and female separate)
Toilets @ 2 m per 1000 peak period market users
Allowance for other specialised uses (e.g auction slab)
Total sales space + ancillary uses in m
Factor of 2.5 to 5 times usable space for roads and parking
Overall site area in m

From the surveys, other physical facilities that the traders may believe are important to locate near the market may have been identified. These might not be accommodated on the market site, but provision can often be negotiated for adjacent sites. Such facilities include:


The purpose of undertaking surveys and making projections is to provide a basis for rational and economic planning. This is important in order to ensure that new and improved market sites and facilities are not over-designed. However, in looking at existing markets it also is important to assess whether it is possible to improve conditions without an intervention requiring new infrastructure. Often, the introduction of new management practices and, particularly, the enforcement of market regulations may provide a more cost-effective means of achieving improvements in market operations without making extensive physical improvements. If the market is subject to strong growth pressures such measures may place excessive pressures on the market users, particularly if there is only limited capacity to raise the efficiency with which the retail space is used. There may also be scope for greater efficiency in transport operations at a market if, for example, more produce is carried on lorries rather than passenger buses.

Contents - Previous - Next