Local and Regional Procurement Learning Alliance
8. Market Data
TRADER SELECTION
Selecting Traders
2
Selecting wholesale and retail traders to interview within
each market
• Objective: track prices of 5 retailers and 5
wholesalers per commodity per market
– In first period, interview up to 10 retailers and up to 10
wholesalers
• There may be some attrition. Initial over-sampling means future
data can be collected by cell phone, even if one or more traders
drop out
– In second period, interview five of each type of trader. The
remaining five traders are in a “reserve pool”
• If one or more of these 5 traders are not available, interview
from the reserve pool
3
Selecting wholesale and retail traders to interview within
each market
– In subsequent periods, rank traders by frequency of
response.
• Interview whichever 5 traders have already responded most
frequently; if one or more is unavailable, interview the nextmost frequent respondent
• Return to the market to sample additional traders when it is not
possible to get at least five responses from the set of ten
initially interviewed
• Small or remote communities may have few traders
– If fewer than ten of a trader type operate in a market,
interview (“census”) all traders of that type in that market
4
First time: Number of trader informants
per market per commodity
Wholesalers
Retailers
Central markets in source country
Up to 10
Up to 10
Central markets in recipient country
Up to 10
Up to 10
Sample of source markets
Up to 10
Up to 10
Sample of recipient markets
Up to 10
Up to 10
Counterfactuals for source markets*
Up to 10
Up to 10
Counterfactuals for recipient markets*
Up to 10
Up to 10
*Pilots without an in-depth component will not need to interview traders in counterfactual markets
5
Repeat interviews: Number of trader
informants per market per commodity
Wholesalers
Retailers
Central markets in source country
5
5
Central markets in recipient country
5
5
Sample of source markets
5
5
Sample of recipient markets
5
5
Counterfactuals for source markets*
5
5
Counterfactuals for recipient markets*
5
5
*Pilots without an in-depth component will not need to interview traders in counterfactual markets
6
Randomly selecting traders to interview
• Use the following process to select traders to
interview for vendors arranged around a market
center:
1. Identify the approximate center of the market
2. For markets arranged along a single street: Flip a coin; you will walk to your
right if you get heads; to your left if you get tails
3. For vendors arranged around a market center: Spin a pen on a flat surface;
you will walk in the direction that the pen-tip points
4. Approximate the time to walk to the edge of the market from the center
5. Cut a sheet of paper into ten slips of the same length. On each slip, write a
single interval: 10%, 20%, 30%, 40% … up to 100%. Put slips into a bag.
Randomly select one slip. Walk this percentage-distance towards the edge
of the market.
6. When you stop walking, select the first trader you reach.
7
Randomly selecting traders to interview cont.
7. Ask that trader whether s/he sells any of the commodities to be monitored.
If not, select a neighboring trader who does sell the commodities.
8. If a trader sells several commodities to be monitored, interview this trader
about all relevant commodities sold.
9. If a wholesaler also sells commodities at retail prices, interview this trader
about both wholesale and retail sales.
10. Record the respondent’s cell phone number(s) for arranging follow up
visits or calls.
After completing each interview, return to the center of the market
and repeat the survey process, beginning with the coin flip / pen
spin, until the correct number of retailers and wholesalers per
commodity is reached.
8
Collecting data from traders over time
•
•
Provide some air-time to all traders to induce them
to have phones ready for follow-up surveys
Ideally, track the same traders each period
–
–
Collect cell phone numbers
Follow up over the phone for regular price collection
•
Replace unavailable trader with one from the
reserve pool
In future periods, follow up with whichever traders
have responded most frequently in previous periods,
drawing from reserve pool to reach 5 traders per
period
•
9
Interviewing Traders
10
Eliciting commodity-specific information
• Fortified or processed commodities may not be sold
directly to household or individual purchasers
– It may not be possible to track retail or wholesale prices for
fortified or processed products
• For blended foods, collect prices of the three main food
ingredients used to produce the fortified products, by cost
– It will generally not be possible to track price series for
supplements or fortificants
• Fortified vegetable oil
– If distributed vegetable oil is branded, track specific brand that
will be distributed
– If fortified oil is not available in markets, track the price for
unfortified oil and note the quality difference
11
Eliciting commodity-specific information
• Characteristics of commodities procured in small lots
from farmer groups, cooperatives or associations may
vary across (or within) groups
– For example, beans varieties may vary by community, agroecological zone, or individual farmer
– If traders do not sell the same commodity that will be
distributed, ask traders for prices on the commodity with
characteristics closest to the distributed commodity
• Showing a sample or photo of the distributed commodity can help
traders identify which characteristics are most similar
– Identify the characteristics that are most common (most
heavily traded) in the central market and use this in the trader
surveys
• Note local names of commodities, and note which are sold in
12
central markets, and which are not
Trader interviews
• For both wholesalers and retailers:
– Need to collect information on commodities identical
to (or most similar to) commodities that will be
distributed / monitored.
– If you are unsure whether the monitored commodity
is available in this market, speak to 3-4 key
informants to determine whether the exact
commodity is available.
– Select only traders selling the commodities of
interest.
• For example, if white maize and yellow maize are
sold and you intend to distribute yellow maize, speak
13
with vendors about yellow maize only.
Trader interviews
• The first interview will take longer than
subsequent interviews
– For both wholesalers and retailers:
• Need to collect some baseline information during the
first interview
– For retailers: need to establish and weigh each
retailer’s standard unit of sale
• Always use your own scale to measure commodities,
not the trader’s scale
• You may need to purchase commodities to measure
them
14
Interviewing retailers for the 1st time
• Measure the standard unit for each retailer
during the first interview
– Zero your scale when empty
• Ask the retailer’s usual unit or volume of sales
for each commodity
– Record the unit of sale by its local name
– Units of sale, such as “cup” or “jerrycan”, vary
widely across traders
• To decrease measurement error, need to know
precise amount each unit holds
15
Interviewing retailers for the 1st time cont.
• Ask for one unit of each commodity
– Weigh the unit
• If it is below two kilograms, ask the trader to give
you an additional unit
– Weigh again
• If the two units are still below two kilograms, ask for
an additional unit, and so on, until the scale reads
at least two kilograms
– Record the number of units required to reach at
least two kilograms.
– Record the scale’s exact reading of the weight
16
Interviewing retailers for the 1st time cont.
• Ask and record the price per unit
• Use a calculator to calculate the price per kg
or liter
• In future interviews, if a retailer changes to a
different unit of sale, measure the new unit
– If collecting information by phone, do not collect
price data for any commodities that have a new
unit of sale until you can measure the new unit.
Instead, ask a reserve pool respondent about
those commodities.
17
Interviewing wholesalers for the 1st time
• A wholesale trader may sell both retail and
wholesale; if so, collect both prices
– On survey, note whether wholesaler is answering
questions about his/her retail or wholesale business
• For retail business, weigh unit of sale and
follow previous approach for retailers
18
Interviewing wholesalers for the 1st time
cont.
• For wholesale business, do not weigh
• Ask the wholesaler’s usual or standard unit or
volume of sales for each commodity
• Record the unit of measurement by name
– Often units of measurement are 90kg or 100kg bags
19
Interviewing wholesalers for the 1st time
cont.
• Record the price per unit
• Use a calculator to calculate the price per kg
– This is an optional step in the field and up to the
discretion of each country office
• In future interviews, verify the wholesaler is
still selling in his/her usual unit of sales.
– Record new unit name and weight/volume, if it
changes
20