11. Socio-economic and engineering surveys

Contents - Previous - Next

Surveys are the foundation of any project and are an essential part of the design process. Some words of caution are required, however, before launching a full-scale survey programme.

Introduction to survey techniques

Like many of the issues covered in this manual, surveys are a vast subject in their own right, which in the case of markets in less-developed countries touches on many other areas of knowledge, including anthropology, statistical analysis, agronomy, agricultural economics and geography. This manual can only cover the most common surveys that will be required for physical planning purposes.

The main difficulty in setting up any survey programme is to obtain the maximum amount of information with the minimum of effort. This may appear to be completely obvious, but there will be strong pressures from many involved in the development process to short cut the surveys, often because of previous bad experiences of over-elaborate and costly surveys which failed to provide the data required.

Only that information should be collected for which there are staff and facilities capable of providing analysis. I or any one day's field work there is at least another day's work in the office. A survey team of, say, six people will generate enough work back in the office to absorb the time of 2 or 3 people for a week, even if they arc using a desk-top computer to analyse the data.

It is essential, therefore, to observe the following principles:

the purpose of the survey should be clearly defined;
there should be a pilot survey before undertaking a full-scale survey;
questionnaires should be short, clear and only cover key questions. Only factual questions should be asked and it should be possible to answer as many of the questions as possible by simply ticking or crossing a box on the form;
surveys should be random, either by selecting every 5th, 10th or so stallholder or lorry driver; or by superimposing a grid over a site plan of the market. They may, however, also be stratified to ensure that all interest groups (such as wholesalers, commission agents and retailers) are covered. To obtain statistically significant results the sample size should be as large as possible, within the constraints of the resources available to collect and analyse the data; and
the survey enumerators must be fully briefed. They should be trained and tested in the field before undertaking the main survey. A team leader should be appointed to coordinate the surveys and to deal with administrative matters such as payment of expenses and provision of seats, food, drinks and umbrellas.

Background reading on surveys are contained in the Bibliography. Particularly useful are those which deal with rapid rural appraisal techniques, including: Holtzman, J.S. (1986. Rapid reconnaissance guidelines for agricultural marketing and food system research in developing countries. Michigan State University, Department of Agricultural Economics, Working Paper No. 30) and McCracken, J.A, Pretty, J.N.& Conway, G.R. (1988. An introduction to rapid rural appraisal for agricultural development. London, International Institute for Environment and Development).

Socio-economic surveys of existing facilities

Which surveys are undertaken will depend on the type of data that is already available and the extent to which it is necessary to supplement it. With existing markets the problem is simpler than with new facilities, where indirect methods of survey may need to be adopted. The basic information needs, however, are similar to those needed for examining an existing market.

Scope of surveys. The types of socio-economic surveys that are needed in the design of wholesale markets are:

quantities and types of goods traded, by season;
types and roles of market users/functionaries;
marketing channels, by season;
employment characteristics of the market, by gender;
management system and operational methods;
rents, tolls and revenues; and
annual turnover and profits.

In addition to the detailed engineering surveys described at the end of this chapter, broad physical surveys are also required of:

the existing market layout and facilities;
space utilization and availability, including stall sizes;
degree of overcrowding/congestion; - utilization of facilities such as cold storage and silos; and
traffic surveys of:
- frequency of trips and departure/arrival times;
- volume of goods transported and modes of transport used; and
- origin and destination of produce.

Survey methodologies. There are three basic methods by which these types of data can be collected:

Observation. Casual (but informed) observation and limited interviews with the market functionaries can provide a wide variety of data. The techniques of rapid rural appraisal arc very valuable in this context. A useful technique is to follow the marketing of a batch of produce from the farm level or local assembly market up to the wholesale market and then follow it right through the market until it is purchased by a retailer or other trader.

Interview surveys. Market users can be interviewed using a random stratified sample and the data can then be expanded to provide an estimate for the overall market. Two methods can be used, sometimes in combination with each other:

by stopping and questioning the drivers of vehicles entering or leaving at the market's check-point or gate house. l his method can be used to establish the weight of deliveries, the vehicle type, its origin or destination and what type of produce is being delivered to or taken out of the market.
by interviewing stallholders or wholesalers on a sample basis. A questionnaire for this method, which has been used for a number of 20 percent sample surveys is shown in Figure 16. The survey enumerator stopped at every fifth stallholder he or she came to. A brief interview was held and a record was taken of the overall volume and types of produce that were expected to be traded that day, the origin of the produce and the expected destination (usually based on the previous day's trade). The survey data can be analysed using computer spreadsheets (the data in the case of a 20 percent sample expanded by 5 times to obtain a 100 percent coverage) to provide the following information:

Figure 16 Interview survey pro-forma for markets in Chiang Mai Province, Thailand

- total daily volumes and type of produce sold; -selling areas of stalls;
- gender and number of employees at each stall; -volume of purchases by channel;
- volume of purchases by origin of produce; and -volume of sales by destination of produce.

Roadside surveys. This is the most reliable and comprehensive method as such surveys record all the produce entering or leaving a city or its main market within a specified time period. This technique is further discussed as a case study in the next section of this chapter.

Case study of roadside survey of commodity flows (Thailand)

The purpose of these surveys (undertaken by FAO Project TCP/THA/8958) was to understand how the Muang Mai market in Chiang Mai operated as a secondary wholesale market for the Northern Region of Thailand.

Scope of surveys. As the market is a private establishment, with only limited intervention by the municipal government, no records were maintained of transactions. It was therefore necessary to undertake surveys to establish the total quantity of fruits and vegetables both entering and leaving Muang Mai on a typical day, including recording their place of origin and the destination to which they arc being sent. This was needed so that the catchment of the market could be defined and produce flows could be quantified by type and mode of transport.

The surveys were undertaken over two full days at the end of the wet season (16/17 August 1990) and one full day in the middle of the dry season (4 May 1991). Eight enumerators were located at key points in the market (see Figure 17) covering all the main entry and exit points. The enumerators were instructed that they should ensure, as far as possible, there should be no duplication. Each vehicle was recorded only twice: when it entered and when it exited the market. The act of unloading a vehicle was considered to be its time of entry and the completion of loading and its departure, its time of exit.

Survey methodology. The enumerators were provided with standard survey forms (see Figure 17), one row of the form- being used for each set of observations (either an exit or an entry of a vehicle). The enumerators were instructed to question the vehicle drivers if there was any doubts about the origin or destination of the produce. The surveys were started promptly at 08:00 am and finished at 16:00 pm, the operating hours of the market. Actual vehicle entry and exit times were recorded and later coded on an hourly basis. Vehicles types were recorded on the form by using a code number:

Head Load/gasket - 0
Push Cart - 1
Bicycle - 2
Rickshaw - 3
Motorbike - 4
Tuk - Tuk (3 wheeler) - 5
Pick-up (1.5 tons) - 6
Saloon Car - 7
4 Wheel Truck (6 tons) - 8
6 Wheel Truck (7.5 tons) - 9
10 Wheel Truck (13 tons) - 10
Other Vehicles - 11

The main commodities that the vehicle was carrying were recorded and the percentage that the particular produce formed of the total load was approximately estimated on the basis of the capacity of the type of vehicle. These data were grouped into six classes of produce:

V1 All leafy vegetables green vegetables
V2 Bulbs, root crops and tubers
V3 Fruit-vegetables (ea. tomatoes), peppers and others
F1 Northern fruits (Iychee, langsat and longan)
F2 Southern fruits (rambutan, mangostecn and durian)
F3 Citrus, melons and others

The total load in metric tonnes was recorded based on the type of vehicle used in carrying the produce. Account was taken of partial loads, such as 1/2 load and empty vehicles were also recorded, as well as those with non-fruit/vegetable loads. The origin or destination of the particular loads were identified as accurately as possible (by province/district/ towns).

Particular care was taken to establish traffic that might be emanating from the Chiang Mai municipal area. The same enumerator did not necessarily record the entry and exit of the same vehicle and vehicle licence plate numbers were also recorded to help establish these data. For analysis purposes the origin/destination (O/D) data of vehicles was

Figure 17 Roadside survey pro-forma and location map of survey points, Muang Mai wholesale market survey, Chiang Mai, Northern Thailand

classified into thirteen groups:

0 within Muang Mai market;
1 within Chiang Mai city;
2 Chiang Mai Province;
3 -11 other provinces in the Northern Region; and
12 outside the Northern Region (Bangkok, the south or the east).

The type of trade being undertaken was also recorded, such as whether it was wholesale or retail. The same vehicle might, for example, bring a wholesale load into the market and leave with retail commodities including non-fruit/vegetable loads.

Analysis of surveys. The coding of the data was first checked against the categories of vehicle types, produce and O/D locations described above. The classified data for each day of the survey was then analysed on a personal computer using a standard spreadsheet programme. Each line in the survey forms was represented by a line in the spreadsheet and was initially sorted on the computer using time as the primary key and origin/destination as the secondary key. The data was then sorted into three basic spreadsheets as follows:

traffic entering the market (ie. destination Muang Mai);
traffic circulating within the market; and
traffic leaving the market (ie. origin Muang Mai).

These three main sets of data were then further sorted, by date and whether incoming or outgoing traffic, into the following categories:

commodity flows and type of produce by origin/destination;
commodity flows by time of entry/exit; and
commodity flows and number of trips by vehicle type/mode.

Results of the wet season survey This section presents an analysis of the two days of the wet season survey; a similar analysis having also been undertaken for the dry season. Table 11.1 summarizes the commodity flows (in tons) in the wet season. There was a good correlation in the figures, both between the two days and between the incoming and outgoing flows. The average daily commodity flow for the rainy season was around 300 tons. The average daily commodity flow for the dry season was around 210 tons, 30 percent less than in the wet season.

Table 11.1 Summary of commodity flows in the wet season

  Commodity Flow (tons):  
Type of Flow August 16 August 17 2 Day Total Av daily
Incoming - coming into market 257.15 347.93 605.08 302.54
Internal - circulating within 24.51 36.93 61.44 30.72
Outgoing - leaving market 297.02 285.26 588.28 291.14
Average 277.08 316.59 - 297.19

Further analysis of the spreadsheet data enabled a number of other summary tables to be prepared, which are given below. The first of these tables (11.2) presents the commodity flow by type of produce. It was clear from the figures that Muang Mai is primarily a vegetable market, with this type of produce representing 90 percent of the trade. The most important source for vegetables during the rainy season was found to be the upland cropping areas in the adjacent districts to Chiang Mail

Table 11.2 Volume of wet season produce by type

Type of Produce Incoming (tons) Outgoing (tons):
Av. daily % Av. daily %
V1 All leafy vegetables/ green vegetables 199.97 66.1 163.87 56.3
V2 Bulbs, root crops and tubers 52.42 8.7 36.81 12.6
V3 Fruit-vegetables, peppers and others 91.36 15.1 66.86 23.0
F1 Northern fruits (Iychee, langsat and longan) 3.30 0.5 2. 10 0.7
F2 Southern fruits (rambutan, mangosteen and durian) 21.72 3.6 5.07 1.7
F3 Citrus, melons/others 36.29 6 0 16.43 5.7
Total 302.54 100.0 291.14 100.0

In an analysis of commodity flows by origin and destination, contained in Table 11.3, a striking difference was found between goods coming into the market and those going out.

Table.3 Commodity flows in the wet season by origin and destination

Origin/Destination by province Incoming (tons) Outgoing (tons)
Av. daily % Av. daily %
Chiang Mai City 6.63 2.2 25.33 8.7
Chiang Mai Province 218.52 72.2 19.92 6.8
Lamphun Province 27.03 8.9 8.00 2.8
Chiang Rai Province 5.05 1.7 32.47 11.2
Phayao Province 0.75 0.2 - 0.0
Uthai Tharu Province 0.75 0.2 - 0.0
Mae Hong Son Province 4.60 1.5 1.30 0.5
Lampang Province - - 0.25 0.1
Tak Province - - 3.00 1.0
Suko Thai Province - - - 0.0
Other Provinces 5.80 1.9 12.40 4.2
Outside the Northern Region 33.69 11.2 188.47 64.7
Total 302.54 100.0 291.14 100.0

Table 11.4 Commodity flows in the wet season by arrival/departure time

Time of entry/exit Incoming (tons): Outgoing (tons)
Av. daily % Av. daily %
08- 09 am 39.95 13.2 18.60 6.4
09- 10 am 41.37 13.7 16.40 5.6
10- 11 am 42.39 14.0 45.59 15.7
11 - 12 am 42.98 14.0 52.20 17.9
12 - 01 pm 50.29 16.6 20.19 6.9
01 - 02 pm 44.67 14.8 45.89 15.8
02 - 03 pm 23.58 7.8 68.46 23.5
03 - 04 pm 13.58 4.5 9.73 3.3
04 - 05 pm 3.75 1.2 14.08 4.9
Total 302.54 100.0 291.14 100.0

The pattern of vehicles entering and leaving the market during the day is shown in Table 11.4. For vehicles entering the market the pattern conformed with what would be expected; a rapid build-up in flow to a mid-day peak, quickly tailing off by late afternoon. For vehicles leaving the market there was a completely different pattern, with two peaks in the late morning and mid afternoon, corresponding broadly to the haulage distances for the produce.

Table 11.5 illustrates an analysis of commodity flow by vehicle type. The main mode for incoming traffic was the light 1 - 1.5 ton pick-up, representing around two thirds of the total, with the remaining one third of the produce being brought by heavier vehicles. For the outgoing traffic the majority of the load (around 75 percent) was carried on 7.5 - 13 ton trucks. Around 15 percent of the total was carried on pick-ups, mostly for short: haul distances within Chiang Mai city and the adjacent districts.

Table 11.5 Commodity flows in the wet season by vehicle type

Vehicle type/mode Incoming (tons) Outgoing (tons)
Av. daily % Av. daily %
Head Load/Basket 0.04 0.0 0.12 0.0
Push Can 2.46 0.8 1.84 0.6
Bicycle - 0.0 3.&6 1.3
Rick straw 0.87 0.3 2.85 1.0
Motorbike - 0.0 1.14 0.4
Tuk - Tuk 0.11 0.0 2.06 0.7
Pick-up (1.5 T.) 198.64 65.7 43.86 15.1
Saloon Car 26.35 8.7 10.46 3.6
4 Wheel Truck (6 T.) 16.08 5.3 0.15 0.1
6 Wheel Truck (7.5 T.) 26.63 8.8 74.71 25.7
10 Wheel Truck (13 T.) 17.58 5.8 141.00 48.4
Other Vehicles 13.80 4.6 9.10 3.1
Total 302.54 100.0 291.14 100.0

Engineering surveys

Apart from the socio-economic studies and traffic surveys described above, engineering surveys will be needed for both existing or proposed market sites . These surveys will typically include:

basic mapping; - detailed topographic and geodetic surveys;
preliminary and detailed site investigations of engineering soils;
an analysis of a site's drainage problems;
structural condition of existing buildings;
surveys and testing of existing services
water supply;
electrical supply;
sewerage disposal;
a review of available construction materials and technologies; and
typical construction costs.

These types of surveys are common to any medium or large-scale development project. However, because market administrators and planners may find that they have to issue terms of reference for topographic surveys and site investigations before design and supervision consultants are appointed (see Chapter 14) further technical details of what they might entail is given below.

Topographic and geodetic surveys. The most essential step in preparing a detailed site plan will be the preparation of a topographic and geodetic survey of the site, at either ]:500 or 1:1,000 scale. The survey should be based on a closed traverse and related to a known datum (defined by its height above mean sea level). To relate levels to this datum may require flying levels to be taken so that that the site levels can be related to a fixed bench mark outside the site.

The survey should define the site's boundaries and pick up all the existing physical features, including buildings and other infrastructure under construction. Spot level should be taken at all breaks in slope, edges and bases of drains and on a 10 metre grid over the entire site. Techniques for undertaking the surveys will vary depending on the local surveyor's skills and availability of equipment. This might range from simple chain traversing, plane tabling and level surveys, through to electronic distance measuring. Advice should be sought from a national survey organization or local surveyor's professional body on the appropriate technical specification and realistic levels of accuracy. Typical standards to be adopted (based on British Standard BS 5606) might be:

accuracy of level values not to exceed 5mm per single sight and 10mm per kilometre relative to a permanent bench mark;
horizontal linear measurements to be accurate within 20mm per 100 metres (applying appropriate slope and tension corrections); and
angular measurements to be accurate to within 20 seconds of arc.

Soils and foundation investigations. Detailed geotechnical investigations are required so that the foundation design of buildings, road pavements and other structures is both safe and economic to construct. The locations of test sites should be established so as to provide this optimum data for building and road design. The site investigations will usually be carried out by a specialist contractor, who will undertake the field survey and laboratory analysis, as well as prepare a comprehensive report.

The scope of surveys will normally include boreholes (typically using an auger and rotary-cum-wash boring rig), dynamic cone penetration tests and California Bearing Ratio (CBR) tests. Disturbed samples are collected from split-spoon samplers and undisturbed samples with open tube samplers. These data will need to taken to a laboratory, where the following analyses and tests should be be undertaken:

standard penetration tests, related to the depth of boreholes and dynamic cone penetration tests, correlated to standard penetration test data.
tests on physical properties:
- natural moisture content;
- Atterberg limits (liquid limit, plastic limit and plasticity index);
- bulk density and specific gravity;
- sieve/grain size analysis (wet method); and
- particle size/hydrometer analysis for clayey soils;
unconsolidated undrained triaxial tests on undisturbed samples, providing stress versus strain curve data;
consolidation tests on undisturbed samples, providing compression index values;
modified Proctor compaction tests on CBR samples;
California bearing ratio (CBR) soaked tests;
chemical tests, including pH and soil sulphate content; and, if appropriate,
soil-cement stabilization tests, using triaxial compression apparatus.

The contractor should be required to prepare a detailed report of the laboratory analysis and to plot the borehole log data to give typical site profiles, indicating the changes in soil types, depth to bed-rock and the groundwater table level. The application of these site investigation results is described in Chapter 14.

Figure 18 Spatial distribution of production areas in Nepal (by season)

Contents - Previous - Next