Content uploaded by Enoch Kikulwe
Author content
All content in this area was uploaded by Enoch Kikulwe on Apr 06, 2023
Content may be subject to copyright.
Content uploaded by Marsy Asindu
Author content
All content in this area was uploaded by Marsy Asindu on Mar 09, 2023
Content may be subject to copyright.
Prices, profit margins
and intermediary market power:
evidence from the matooke value
chain in Uganda
Rob Kuijpers, Esther Smits and Cedric Steijn
Sustainable Economic Development, KIT Royal Tropical Institute,
Amsterdam, The Netherlands
Nasser Mulumba
Alliance of Bioversity International and CIAT, Kampala, Uganda
Marsy Asindu
International Livestock Research Institute, Hanoi, Viet Nam and
Kiel University, Kiel, Germany
Froukje Kruijssen
Sustainable Economic Development, KIT Royal Tropical Institute,
Amsterdam, The Netherlands, and
Enoch Mutebi Kikulwe
Bioversity International, Maccarese, Italy
Abstract
Purpose –There is widespread belief that intermediaries in African agri-food value chains have
disproportionate market power. In this paper, the authors examine this belief by uncovering the purchasing
and selling prices, costs and profit margins by farmers, intermediaries and retailers in the matooke (cooking
banana) value chain in Uganda, and by analysing the prevailing value chain and market structures, seasonal
entry and exit dynamics and the trading relationships in the chain.
Design/methodology/approach –Data for this study were collected along the trading routes from the main
matooke producing districts in South-West Uganda (Kabarole, Bunyangabo, Bushenyi, Isingiro and Mbarara)
to the main urban markets around the capital Kampala. A structured survey was administered with 383
producers, 172 collectors and wholesalers and 71 retailers. In addition, key informant interviews and focus
group discussions were held.
Findings –The authors find that price mark-ups by intermediaries (selling prices minus purchasing prices)
vary with the type of intermediary, season and location but generally reflect the costs of moving matooke down
the value chain to the urban consumer. The authors do not find evidence for disproportionate market power
among the intermediaries in the chain. Intermediaries enter and exit the market in peak and off-peak season,
such that profits are kept in check. This seasonality does imply a small shift in market power in favour of
farmers in off-peak season and in favour of intermediaries in the peak season.
Research limitations/implications –The investigation concentrated on an important and relatively
homogenous staple crop along its main trade route. More remote areas, where there is less of an abundance of
matooke, might still be characterised by local monopsonies where intermediaries have more market power due to
high search and transport costs. Similarly, (local) monopsonies might exist for products for which there is a smaller
market (segment), for products with a stronger seasonal variation in supply and for more perishable products.
Market power
in the matooke
value chain
The authors acknowledge research funding from the CGIAR Research Program on Policies, Institutions
and Markets (PIM) under Flagship 3: Inclusive and Efficient Value Chains. The authors are grateful to
the participants in the research for providing information and the enumerators involved in the data
collection. The authors thank Dr. Hector Chavez for providing support with mapping of survey
locations.
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/2044-0839.htm
Received 2 June 2022
Revised 7 December 2022
Accepted 15 December 2022
Journal of Agribusiness in
Developing and Emerging
Economies
© Emerald Publishing Limited
2044-0839
DOI 10.1108/JADEE-06-2022-0105
Originality/value –While there is an important literature on the role of intermediaries in African agri-food value
chains, the evidence on intermediary market power is scant. Beliefs on intermediary market power are largely
based on anecdotal evidence from farmers or inferred from observed prices or market structures. The paper
contributes in addressing this important knowledge gap by studying the matooke value chain in Uganda.
Keywords Market power, Profit margins, Intermediaries, Competition, Matooke
Paper type Research paper
Introduction
Food and agricultural value chains are widely acknowledged to provide major opportunities
for economic development and poverty reduction in Africa. Yet, there are also concerns that
agri-food value chains are characterized by a concentration of power and profits in certain
nodes of the chain at the detriment of farmer and consumer welfare. In particular, there is
widespread belief among farmers, NGOs, researchers and policy makers that intermediaries
have too much market power, earn excessive profit margins and drive a price wedge between
farmers and consumers (e.g. Barrett et al., 2020;Chau et al., 2016;Mitra et al., 2018). This has
motivated policy makers to intervene in markets to improve the terms offered to farmers, for
example, by investing in information technology, market regulations, state-run marketing
boards, Fair Trade certification and support to farmer organisations.
The evidence on the market power of intermediaries in African value chains is,
however, scant [1]. While there is an important literature on the role of intermediaries in
developing countries (e.g. Fafchamps and Gabre-Madhin, 2001;Gabre-Madhin, 2001;
Fafchamps et al., 2005;Chau et al., 2016;Abebe et al., 2016;Minten et al., 2017), there is
particularly little empirical evidence on their market power. Key exceptions include recent
empirical papers on Asian agri-food value chains suggesting there is a degree of market
power exercised by intermediaries (e.g. Mitra et al., 2018;Kopp and Sexton, 2021;Rahman
et al., 2020;Kopp and Mishra, 2022). Most of the robust evidence on intermediary market
power is, however, from middle- and high-income countries (see, e.g. Xhoxhi et al., 2020). A
recent in-depth review of this literature did not yield evidence of “systematic and large
abuses of market power by other players at the expense of farmers”(Deconinck, 2021, p. 4).
Beliefs and statements on intermediary market power in African agri-food value chains
are, so far, largely based on anecdotal evidence (e.g. from farmers) or inferred from
observed prices along the value chain or from the observed market structures, with high
concentration at the intermediary level.
Our paper contributes in addressing this important knowledge gap by uncovering the profit
marginsearnedbyintermediariesalongthematooke value chain in Uganda and by examining
the prevailing value chain structure, market entry- and exit-dynamics and trading relationships.
Matooke is the most important staple crop in Uganda (Haggblade and Dewina, 2010) and provides
major employment for farmers, intermediaries and workers in the chain (Steijn et al., 2022).
Matooke farmers have accused wholesalers (lorry traders) of paying unfair prices, while
wholesalers, in turn, argue that they have high operating costs and not always make a profit
(Zijlstra, 2015). Earlier research on the matooke value chain has focused on the composition of
the chain (Nalunga et al., 2015), postharvest losses (Kikulwe et al., 2018) and gender relations
(Rietveld et al., 2016). Two studies provide some insights on prices and profit margins. Kalule
and Kyanjo (2013) studied the marketing margins and efficiency at the retail node of the
matooke chain in Uganda but do not provide a full overview of the entire value chain.
Nalunga et al. (2015) present the price structure in the full value chain and do find that the
highest price mark-ups (computed as selling price minus purchasing price) are realized at
the wholesaler node of the value chain [2]. However, they do not provide information on costs
incurred and therefore cannot present information on profit margins.
Using a stacked value chain survey design, we provide a descriptive analysis
of purchasing and selling prices, operating costs, margins and profits at the different
JADEE
nodes of the value chain: at the level of farmers, intermediaries (collectors, brokers
and wholesalers) and retailers. This allows analysing the extent to which price mark-ups by
intermediaries reflect costs for moving the product from the farmer to the consumer
or whether the profits earned in the middle segment are indeed excessive. We do
this separately for the peak and off-peak seasons. This is complemented by information from
semi-structured interviews and focus group discussions with farmers, intermediaries and
retailers to better understand the value chain and market structure, entry and exit dynamics,
the role of seasonality and relationships between different value chain actors.
Our analysis suggests that the large differences in mark-ups (the spread between selling
price and purchasing price) observed between different actors in the value chain reflect, to an
important degree, differences in operational and wage costs. This is particularly apparent at
the wholesale level where the relatively high mark-ups appear to be necessary to cover for the
cost of trading, particularly in the off-peak season when volumes are lower. The hypothesis
that wholesalers pay unfair prices and earn the highest per-unit profit margins is thus not
supported by our analysis.
We also observe an increase in price mark-up from the peak to the off-peak season for all
actors, which reflects both a shift in underlying costs as well as a potential shift in relative
market power. For collectors and wholesalers this increase in mark-up just about offsets the
lower trade volumes and the higher costs per quantity traded, such that profit margins stay
constant throughout the year. Retailers and farmers (those that stay in business), however,
experience a significant increase in per-unit profit margins in off-peak season, potentially
reflecting a relative shift in market power in their favour.
Finally, although women are participating in all segments of the value chain, we find that
women active in trading and retailing earn significantly lower profits than their male
counterparts. This seems to be a result of lower trade volumes rather than differences in per-
unit profit margins and points to other underlying barriers for women, such as (but not
limited to) a lack of access to capital and other resources to operate at a larger scale.
The remainder of this paper is structured as follows. We first present the methodology.
We then provide a short overview of the matooke value chain, including description of the actors,
the characteristicsof the chain and the market structure. Next, we provide quantitative estimates
of price mark-ups, per-unit profit margins and total profits by the different actors in the chain
and analyse the role of costs, seasonality and gender. We conclude the paper by summarizing
our main results and limitations and by discussing recommendations for future research.
Methodology
In a spot-market exchange, a buyer has market power over a seller if the negotiated sales price is
below the price level that would emerge in a competitive market. This (local) market power thus
arises from imperfect competition caused by, for example, high costs of finding buyers,
information asymmetry or a small number of alternative buyers active in the market. Market
power is thus a latent variable that may be inferred from observable variables related to the
exchange.
Broadly, three types of indicators have been used to empirically infer market power in agri-
food chains: “structure”,“conduct”and “performance”indicators (Bonanno et al.,2018)[3].
Market structure indicators measure market concentration (e.g. the combined sales of the four
largest firms), assuming market power arises from such concentration. Conduct indicators
assess whether companies behave, in terms of price setting, as if they are in a competitive market
(e.g. paying a price equal to the input’s marginal value product). There is an emerging
literature—new empirical industrial organisation—that uses this approach (see, e.g. Rahman
et al., 2020). The third way to infer market power is to assess performance of firms, ideally by
assessing the realized profit margins—with higher profit margins assumed to be an indication
Market power
in the matooke
value chain
for higher market power. The advantage of such performance indicator is that it relies on less
strict assumptions regarding the market structure and—unlike the conduct indicators—can
account for fixed costs (Deconinck, 2021). In this paper, we primarily estimate performance
indicators to infer market power, but we complement this with an assessment of the structure of
the matooke sector based on qualitative data.
Data collection
Data for this study were collected along the trading route from the main matooke producing
districts in the South-West of Uganda to the main urban markets around the capital Kampala
(see Figure 1). Data were collected from all segments along the matooke value chain: farmers,
intermediaries and retailers. Quantitative and qualitative data from the farmer and
intermediary segment of the value chain were collected in October 2019 from the Western
region of Uganda, which produces two-thirds of the national volume of matooke production
(UBOS, 2020a). Within the Western region, we purposefully selected the five districts with the
highest matooke production (producing about 100,000 MT and above each): Kabarole,
Bunyangabo, Bushenyi, Isingiro and Mbarara (UBOS, 2020b). From these districts matooke
is traded to the urban markets. Data on the retail segment of the value chain were collected
between July and August 2020 from key destination markets in the urban districts of
Kampala, Jinja and Wakiso. The surveys for the three actors capture similar types of
information, including socio-economic and demographic characteristics, marketing channels
used, volumes of matooke produced, bought and sold, purchasing and selling prices, costs,
asset investments, numbers of workers employed and type of employment contracts.
Quantitative data were collected among 383 producers, 172 intermediaries and 71
retailers (Table 1)[4]. Farmers were sampled using a three-stage random sampling
procedure. First, the top eight banana producing sub-counties for each district were
identified and a random sample of two sub-counties from each district was taken. Second, a
Figure 1.
Map with sampled
production and
consumer market
districts and the two
main matooke trading
routes
JADEE
list of villages in the sub-counties was constructed and two villages were randomly selected
from this list. Third, lists of farmers were obtained with the help of the local council
authorities and random samples for interviews were drawn based on these lists, selecting
43 farmers per village. The intermediaries and retailers on the other hand were sampled
using a two-step approach. Markets were selected based on average volume traded in each
market, selecting the largest markets. From each market, a list of intermediaries and
retailers active on the market was made together with the market masters/chairpersons.
Intermediaries and retailers to be interviewed were then randomly sampled from these lists.
The number of intermediaries and retailers interviewed is proportional to total number in
the respective markets.
Qualitative data were collected in the same districts and markets as the quantitative survey
data, using Key Informant Interviews (KIIs) and Focus Groups Discussions (FGDs) with various
actors in the matooke chain, including farmers (15 FGDs), intermediaries (6 KIIs), labourers
(6 FGDs), cooperatives (4 KIIs) and a matooke processor (1 KII). Farmer FGDs were conducted in
three disaggregated groups: men, women and youth (<35 years old). These data were used to
support, verify and contextualise the findings of the quantitative analysis.
Calculation of mark-ups and per-unit profit margins
Based on information provided by the different actors along the chain, we proceed to compute
mark-ups, per-unit profit margins and profits for each actor. One challenge is that bananas in
the markets are traded in bunches that are classified as small, medium and large. To obtain
standard units and ensure comparability, we convert the bunches to kilogram-equivalent by
randomly sampling and weighing forty typical bunches of each size from each of the markets
visited within the study districts. The average weight for each size is used as the conversion
rate for the respective sizes for each district (see Table 2).
Another challenge is that value chain actors do not keep records of yearly volumes of
matooke produced or traded and that the matooke harvesting season is not concentrated in one
District
Farmers Collectors Wholesalers Retailers
Men Women Men Women Men Women Men Women
Bushenyi 39 28 12 0 3 4 ––
Bunyangabu 28 41 8 1 7 6 ––
Kabarole 50 47 26 5 7 8 1 1
Isingiro 47 42 14 1 7 3 ––
Mbarara 24 37 40 0 10 8 ––
Kampala –––– 01322
Jinja –––– 01421
Wakiso ––––––118
Total 188 195 100 7 34 31 9 62
District Small-sized bunch Medium-sized bunch Large-sized bunch
Mbarara 21.3 31.6 54.3
Isingiro 17.1 25.8 35.3
Kabarole 13.3 26 33.8
Bushenyi 21.3 29.6 43.4
Bunyangabu 19.3 34.8 42.0
Note(s): As no data was available for Bushenyi, the average bunch weight of other districts was used
Source(s): Own data collected by weighing 40 bunches of bananas of each size in each market
Table 1.
Sample size by district,
actor and gender
Table 2.
Average weight (kg) of
small, medium and
large sized bunches in
each of the districts
Market power
in the matooke
value chain
or two short periods (as is the case with cereals, for example). This makes it difficult to estimate
the total volumes of matooke produced/traded per year. Figure 2 shows the number of farmers
reported to have harvested matooke in a given month. In every month of the year, there is some
matooke production, but the peak production season runs from June to August. Therefore,
farmers were asked about production and trade volumes and prices in an “average week”during
the peak season and another average week during the off-peak season (September to November).
We separately analyse the peak and off-peak season, because we expect this seasonality in
volume to affect the supply chain in a number of ways, including prices, the number of actors
active in the chain, market power and the marketing channels of the farmers.
We constructed our key variables as outlined in Table 3. The intermediary and retailer
surveys provide information on the operational costs incurred in the month before the
interview [5]. Operational costs include costs for building rent, equipment rent, fuel and
electricity, loading and unloading services, commissions, losses, theft, repair, packaging,
overhead, asset depreciation and taxes. We assume non-linear asset depreciation of 5% per
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Share of farmers reporting matooke harvest
Source(s): Own data
Concept Notation and definition
Selling price (large, medium, small) PL;PM;PS
Quantity sold (large, medium, small) QL;QM;QS
Purchasing price (large, medium, small) WL;WM;WS
Revenue Revenue ¼PL*QLþPM*QMþPS*QS
Procurement costs CostsBuying ¼WL*QLþWM*QMþWS*QS
Conversion factor (bunch to kg) FL;FM;FS
Quantity sold (kg) Qkg ¼FL*QLþFM*QMþFS*QS
Selling price (kg) PKG ¼Revenue=Qkg
Purchasing price (kg) WKG ¼CostsBuying=Qkg
Mark-up Markup ¼PKG −WKG
Profit
π
¼Revenue −CostsBuying −CostsOperational −Costswages
Profit Margin Margin ¼
π
Qkg ¼PKG −WKG −
CostsOperational
Qkg
−
Costswages
Qkg
Figure 2.
Share of farmers
reporting harvest by
month (%)
Table 3.
Concepts, notation and
definitions
JADEE
year of the total value of assets. Assets included are buildings, vehicles and equipment
(e.g. scales). We assume the obtained information on operational costs is representative for
both the peak and off-peak season. The survey also captures the costs incurred for wages for
permanent labour (skilled and unskilled), seasonal and part-time employees.
The matooke value chain in Uganda
Matooke, formally referred to as East African Highland banana, is an important source of
food security and livelihoods in Uganda. Ugandans consume, on average, more than
200 grams matooke per day [6]. Matooke is grown by 47% of agricultural households in
Uganda, who together produce 6.5 million mega ton (UBOS, 2020b). In addition to farmers, the
matooke chain provides jobs to other self-employed actors, ranging from collectors using
bicycles and small-scale informal retailers to larger collectors who use trucks, wholesalers
and wage labourers both on and off-farm (Steijn et al., 2022).
Matooke is produced by smallholder farmers that depend on the crop for income and as a
source of food (Ariho et al., 2015). There is little organisation among matooke farmers;
previous studies by Ariho et al. (2015) and Nalunga et al. (2015) found that only 2–3% sold
their produce collectively. The vast majority of farmers sell their matooke to small traders,
such as bicycle, motorbike and small truck collectors. These collectors transport and sell the
matooke to brokers or directly to wholesalers at markets or collection points. Larger
commercial farmers may sell directly to brokers or wholesalers. The brokers bulk the
matooke from the small collectors before selling it to wholesalers. Wholesalers are large
traders from urban areas that transport matooke to markets in Kampala, Jinja or Entebbe
using large trucks. Sometimes brokers are employed by wholesalers. There is also some
confusion about the two terms and they are sometimes used interchangeably in practice. For
this reason we lump brokers and wholesalers as one category in the remainder of this paper.
Retailers in urban markets purchase matooke from the wholesalers before selling it to
consumers, hotels or restaurants.
Most matooke is commonly prepared in thehousehold by steaming or boiling the fingers (i.e.
individual banana fruits) and mashing them. Only a small part of the matooke is exported
(approximately 10%) or processed into wine or flour by small-scale processing firms (Ariho et al.,
2015;Nalunga et al.,2015). The analysis in this paper is focussed on the major value chain
segment of fresh (unprocessed) matooke for the domestic market (illustrated by Figure 3).
There is free and frequent entry and exit of actors during the year. The harvest and number of
farmers active on the market is highest in peak seasonbutsignificantlydrops in off-peak season
(see Figure 2). During the off-peak season, also intermediaries exit the market as the market volume
shrinks and high quality matooke is difficult to find. At this time of the year, intermediaries come to
the farmers and almost all sales take place at the farm-gate. During the peak-season, even though
more intermediaries are becoming active, farmers experience increasing difficulty to find buyers.
As a result some farmers enter the intermediary node of the value chain—as “collectors”—to
transport the matooke to the market by bike or motorcycle themselves.
The qualitative data suggests that this seasonality is also reflected in the type of trading
relationships. To secure trade, both intermediaries and farmers are found to offer credit
services. In the peak season, farmers sell matooke on credit and bear the risk of non-payment
to secure a sales outlet. In the off-peak season, the tables turn and intermediaries are found to
pre-finance matooke, or provide in-kind payments to secure supply.
Descriptive statistics
Table 4 presents the descriptive statistics of the farmer sample disaggregated by sex of the
respondent. Matooke is an important crop for these farmers. Farmers, on average, report to
Market power
in the matooke
value chain
have been in matooke production for about two decades and have more than half of their land
under matooke cultivation. Most farmers produce matooke for either a pure food security
objective or for both food security and income objectives. Only a minority of less than 10%
produces matooke for a pure income objective. Despite the food security focus, the majority of
farmers does sell a share of their harvest. This most commonly happens at the farm gate.
Table 5 presents descriptive statistics for the full sample of intermediaries and separately
for male and female intermediaries. It shows women traders are different from men in at least
three aspects. First, trading matooke is a popular occupation for women before they are
married. Indeed, the majority of women traders in our sample are unmarried (54%). Second,
women are more likely to have used credit to run their business. Third, women are more likely
to use trading as a part-time occupation. They, on average, trade four days per week, while
men trade five days per week.
Results
Table 6 shows average purchasing and selling prices per kilogram (kg) for each of the value
chain actors, as well as the mark-up that occurs as matooke moves down the value chain for
each of the intermediary types. The mark-up is defined as the difference between the selling
price and the purchasing price.
The prices obtained from the farmer and intermediary surveys correspond relatively well
with each other: the farmers’selling prices corresponds with collectors’purchasing prices,
Figure 3.
Simplified model of the
matooke value chain in
Uganda
JADEE
and collectors’selling prices matches the wholesalers’purchasing prices. There is slightly
more discrepancy, however, between the wholesaler selling prices and the retailer purchasing
prices. We would expect these are similar as the wholesaler sells the matooke to the retailer.
The discrepancy in selling and purchasing prices can potentially be explained by the
different period in which the retailer survey was administered (July and August 2020)
compared to the intermediary survey (October 2019). Indeed, monthly price data from 2020 to
2019 from the Uganda Price Bulletin (see Figure 4) suggest retail prices were indeed
significantly lower in 2020 than in 2019, particularly in the months preceding the retailer
survey—February to June 2020 [7]. One factor that might have had a negative effect on prices
in 2020 is Covid-19, which might have induced consumers to buy dried, long storage, staple
products instead of fresh, perishable, matooke.
Full sample
(N 5383)
Male farmer
(n 5195)
Female farmer
(n 5188) T or
χ
2
Age of respondent 48.41 (16.14) 50.82 (15.85) 45.91 (16.09) 3.01***
Respondent is the household
head (%)
82.77 100.00 64.89 82.71***
Education level (completed
years)
5.75 (4.09) 6.59 (4.17) 4.87 (3.84) 4.20***
Years in matooke production 20.94 (14.01) 22.33 (14.11) 19.49 (13.79) 1.87**
Has off farm income (%) 41.78 42.05 41.49 0.01
Accessed market information
(%)
30.29 38.97 21.28 14.19***
Matooke intercropped (%) 47.52 41.03 54.26 6.72**
Cultivable land (acres) 3.92 (4.77) 4.70 (5.55) 3.11 (3.64) 3.30***
Land under banana (acres) 2.02 (2.69) 2.42 (3.05) 1.61 (2.20) 2.96***
Main production objectives
Mainly food security (%) 22.45 17.95 27.13 4.63**
Both food security and income
(%)
69.19 70.77 67.55 0.68
Mainly income (%) 8.36 11.28 5.32 4.44**
Sale points
Farm gate (%) 80.42 79.49 81.38 0.22
Village market (%) 9.14 12.82 5.32 6.49**
Collection point (%) 1.31 1.03 1.60 0.24
Never sells (%) 13.84 11.28 16.49 2.18
Note(s): Standard errors in parentheses. ***p< 0.01, **p< 0.05, *p< 0.1
Full sample Male intermediaries Female intermediaries T or
χ
2
Age 36.35 (10.13) 36.4 (10.50) 36.16 (8.77) 0.126
Education level (years) 6.59 (3.48) 6.40 (3.49) 7.29 (3.42) 1.391
Married (%) 79.65 88.89 45.95 33.04***
Member of an association (%) 25.00 26.67 18.92 0.93
Accessed credit (%) 39.45 34.52 56.00 3.72*
Years in matooke business 9.39 (7.84) 9.76 (7.86) 8.07 (7.69) 1.162
Days per week tradingmatooke 4.87 (2.02) 5.09 (1.93) 4.08 (2.17) 2.745***
N 172 135 37
Note(s): Standard errors in parentheses. ***p< 0.01, **p< 0.05, *p< 0.1
Table 4.
Descriptive statistics
(mean and standard
deviation) for men and
women farmers
Table 5.
Descriptive statistics
(mean and standard
deviation) for male and
female intermediaries
Market power
in the matooke
value chain
Peak season Off-peak season
Purchasing price Selling price Mark-up Mark-up (%)* Purchasing price Selling price Mark-up Mark-up (%)
Farmers Average –132 –– – 350 ––
Mbarara –95 –– – 256 ––
Isingiro –150 –– – 421 ––
Kabarole –152 –– – 364 ––
Bushenyi –112 –– – 338 ––
Bunyangabu –133 –– – 313 ––
Collectors Average 138 219 81 59% 320 447 127 40%
Mbarara 119 198 79 66% 285 383 98 34%
Isingiro 168 267 99 59% 358 476 118 33%
Kabarole 153 232 79 52% 390 577 187 48%
Bushenyi 145 211 66 46% 302 423 121 40%
Bunyangabu 135 207 72 53% 259 371 112 43%
Wholesalers Average 241 502 261 108% 441 851 410 93%
Mbarara 201 381 180 90% 341 828 487 143%
Isingiro 454 1,038 584 129% 680 1,309 629 93%
Kabarole 215 423 208 97% 454 736 282 62%
Bushenyi 179 429 250 140% 411 671 260 63%
Bunyangabu 196 431 235 120% 362 800 438 121%
Retailers Average 285 495 210 74% 608 927 319 52%
Kampala 281 511 230 82% 576 851 275 48%
Jinja 307 510 203 66% 737 1,054 317 43%
Wakiso 266 456 190 71% 496 892 396 80%
Note(s): Wholesalers include brokers as in practices these are often employed by wholesalers. * Mark-up as a percentage is the mark-up (selling price minus purchasing
price) divided by the purchasing price
Source(s): Own survey data
Table 6.
Average purchasing
and selling prices
(UGX/kg) and mark-up
by actor, season and
district
JADEE
Based on the results in Table 6, we make three additional observations. First, the mark-ups in
the wholesale and retail nodes of the value chain—261 and 210 UGX/kg in peak season and
410 and 319 UGX/kg in the off-peak season, respectively—are significantly higher than in the
collector node—where the mark-up is 81 UGX/kg in peak season and 127 UGX/kg in off-peak
season.
Second, there are important differences between peak and off-peak seasons. In line with
the law of demand and supply, farmers can sell their produce for a significantly higher price
during off-peak season (350 UGX/kg)—a time with low supply—than during peak season
(132 UGX/kg)—a time with high supply of matooke. This causes prices in the rest of the value
chain, up to the consumer level, also to be significantly higher in off-peak season, as well as
their mark-ups. On average, collectors, wholesalers and retailers realize an absolute mark-up
which is, respectively, 47, 160 and 110 UGX/kg higher in off-peak season.
Third, farmers receive 26% of the consumer expenditure on matooke in the peak season
and about 38% in the off peak season. This means that the remaining 74 and 62%, in peak
and off-peak season, respectively, goes to intermediaries and retailers in the value chain. This
Figure 4.
Matooke retail prices in
Kampala and Mbarara
for 2019, 2020, and a
5 years average
Market power
in the matooke
value chain
is in line with the global average estimated farm share of food expenditure in the range of 16–
38% (Yi et al., 2021).
Profit margins and market power
To assess whether these mark-ups are indeed excessive and a potential result of market
power, we need to take into account the operating and wage costs. Table 7 presents the
estimations of operational costs, wage costs, per-unit profit margins, volumes and profits in
an average week during peak season and off-peak season, respectively. Figure 5 shows the
distribution of the calculated per-unit profit margins by business type and season.
Table 7 suggests that the variation in mark-ups at the intermediary and retail level reflect the
variation in operational and wage costs. This is most apparent at the wholesale level where the
relatively high mark-ups are necessary to cover the cost of trading, particularly in the off-peak
season. Wholesalers incur by far the highest transport costs as a result of the large distances
from the producing districts to urban consumer markets. Costs include wages for truck drivers
and for on-loading and off-loading, fuel and asset depreciation or rent for vehicles and
Farmers Collectors Wholesalers Retailers
Peak Off-peak Peak Off-peak Peak Off-peak Peak Off-peak
Selling price (UGX/kg) 133 350 218 449 509 855 497 927
Purchasing price (UGX/kg) ––139 322 249 447 285 604
Operational costs (UGX/kg) 9 20 38 81 239 409 98 86
Wage costs (UGX/kg) 19 41 7 15 106 192 68 79
Per-unit margins (UGX/kg) 105 289 33 30 38 51 92 164
Profit rate 79% 83% 15% 7% 7% 6% 18% 18%
Volume (KG) 557 271 5,703 3,321 11,932 7,102 2,752 2,361
Weekly profit
(UGX 31,000)
57 76 220 200 865 767 159 343
N 247 247 95 95 38 38 62 62
Table 7.
Prices, costs, per-unit
profit margins,
volumes and weekly
profit for average week
in peak and off-peak
seasons
Figure 5.
Distribution of per-unit
profit margins by
business type and
season
JADEE
warehousing. In fact, per-unit profit margins earned by wholesalers, in both absolute and
percentage terms, are lower than those earned by farmers or retailers. The hypothesis that
wholesalers pay unfair prices and earn the highest margins (see, e.g. Warsanga and Kilimanjaro,
2014;Nalunga et al., 2015;Zijlstra, 2015) is thus not supported by our analysis. It is tr ue, however,
that wholesalers, on average, make most profit—as argued by Nalunga et al. (2015)—but this
seems to be a result of the size of these businesses and the volumes they deal with, not because
they pay unfair or uncompetitive prices to farmers and collectors. The qualitative data also does
not provide any evidence for collusion among wholesalers and other intermediaries. There
seems to be a large numbers of competitors, particularly in the off-peak season, such that
alternative buyers willing to pay market prices can be found.
The higher mark-ups in the off-peak season charged by collectors and wholesalers,
compensates for the lower trade volumes and the higher operational and wage costs per
kilogram traded. Indeed, a t-test suggests that the difference in per-unit profit margins is not
significantly different from zero between the peak and off-peak season for the collectors
(tvalue 50.15) and wholesalers (tvalue 50.21). These higher mark-ups in off-peak season
thus do not result in higher per-unit profit margins.
Retailers and farmers, those who continue to be active, seem to benefit from the low supply
during the off-peak season. Despite the lower traded volumes, retailers (tvalue 52.3) and
farmers (tvalue 518.9) have significantly higher per-unit profit margins in the off-peak season
than in the peak season. They can offset a dip in volumes by increasing selling prices and
margins. This apparent shift in market power from intermediaries towards retailers and farmers
in the off-peak is also perceived by intermediaries and farmers themselves, as expressed in
interviews and FGDs. For instance, one farmer in Kyakunthu, explained: “During the peak
season, you have to reduce the price to encourage buyers to buy from you. During the off-peak
season, there are plenty of traders so you do not have to do a lot to impress them”.Another
farmer, from Kitonzi A, said: “When they [matoke bunches] are in abundance [during peak
season], you have to take them to the road or sell them to a bike trader. Nobody takes time to
come here and the prices plummet. Sometimes you even abandon the matooke in the market to
avoid transporting them back. This is the time when we mostly sell on credit to still be able to sell
some.’’ Another factor potentially improving the market power of farmers in the off-peak season,
identified in the farmer FGDs, is that matooke is also less perishable during the off-peak season
(the ripening process is slower). This gives farmers more time to wait for the best price.
Despite farmers obtaining per-unit profit margins comparable to other actors in peak
season and higher margins in the off-peak season, farming seems to be the least profitable
activity in the chain in absolute terms. This is caused by the low volumes being sold by
farmers. This is in line with findings by Nalunga et al. (2015),Warsanga and Kilimanjaro
(2014) and Komarek (2010) who found that in East African matooke value chains, the
production node is the “least profitable”. However, our data cannot offer a complete picture of
profitability. A complete assessment of profitability should consider time investments
of household labour, opportunity costs, own consumption of production and the cultivation of
other crops besides matooke. We have not collected data on these aspects.
Lastly, although collectors obtain the lowest per-unit profit margins in the value chain,
they can sustain a stable profit of UGX 200,000-220,000 per week, on average. There is,
however, much variation between collectors that is not reflected in these estimates (see
Table 8). Small truck collectors, in terms of volumes, are more similar to wholesalers than to
collectors by bicycle and motorcycle. Due to their lower per-unit profit margins, they are,
however, not able to obtain the same profits as wholesalers. Motorcycle collectors are
surprisingly similar to bicycle collectors in terms of profits. While they are able to move a
relatively larger volume of matooke, particularly in the peak season, they also have higher
operational costs and lower per-unit profit margins, which offsets these higher volumes.
Market power
in the matooke
value chain
Difference in profits and per-unit profit margins for women
We now assess whether women who participate in the trading and retailing segments benefit
from participation to the same extent as men. Some business activities are male-dominated while
others are female-dominated (see Table 1). Collectors are predominantly men—only 5% are
women—whileretailersarepredominantlywomen—more than 90%. The wholesale segment is
found to have the most equal participation of men (53%) and women (47%). To assess benefits
from participation in the value chain, we thus have to control for the type of business in which
the person is engaged. We do this by regressing profits, per-unit profit margins and volumes on
the gender of the respondent while controlling for the type of business.
The results, presented in Table 9, suggest that women earn lower profits than men in the
same business activity. Women earn, on average, 560,437 UGX less than men in an average
week in peak season and 368,009 UGX less in off-peak season. This does not seem to be the
result of lower per-unit profit margins (and potentially a difference in market power)—these
are found to be like the per-unit profit margins obtained by men—but seem to be the result of
lower volumes traded. In fact, women on average trade about 3,000 kg less than men per week
in peak season and about 2000 kg less per week than men in off-peak season. Potentially, this
is a result of the fact that women are more likely to work part-time, as shown earlier. Another
potential reason is lower availability of cash and investment capital, leading to less access to
Peak season Off-peak season
Bicycle Motor-cycle Small truck Bicycle
Motor-
cycle Small truck
Selling price (UGX/kg) 212 219 265 441 445 526
Purchasing price (UGX/kg) 136 144 158 318 320 365
Operational costs (UGX/kg) 269 66 403 52 157 1,027
Wage costs (UGX/kg) 4 2 46 10 2 96
Per-unit margins 484 8 19 61 33 42
Volume (kg) 4,657 6,927 10,830 3,040 3,091 6,233
Weekly profit (UGX 31,000) 235 165 248 268 42 83
N 65 22 8 65 22 8
(1) (2) (3) (4) (5) (6)
Profit (peak)
Profit
(off-peak)
Per-unit
profit
margin
(peak)
Per-unit
profit
margin
(off-peak)
Volume
(peak)
Volume
(off-peak)
Business owner
is female
560,437* 368,252 47 37 3,095** 1,949**
(305,289) (286,364) (35) (56) (1,284) (749)
Retailer 416,356 456,129* 99*** 103* 318 699
(270,704) (261,612) (35) (54) (1,243) (782)
Wholesaler 881,802** 721,836** 24 6 7,533*** 4,602***
(346,775) (356,093) (33) (62) (1,398) (849)
Constant 249,091*** 219,784*** 36*** 28 5,866*** 3,423***
(59,810) (74,486) (10) (19) (560) (386)
Observations 195 195 195 194 195 195
R
2
0.106 0.060 0.045 0.072 0.287 0.207
Note(s): Standard errors in parentheses. ***p< 0.01, **p< 0.05, *p< 0.1
Table 8.
Costs, per-unit profit
margins, volumes and
profits for peak and off
peak season by type of
collector
Table 9.
Regression results for
the relation between
gender and profits, per-
unit profit margins and
volumes
JADEE
transportation and storage facilities. Indeed, we have shown earlier that women
intermediaries are more likely to use credit than men.
Conclusion
We investigated whether intermediaries in the Ugandan matooke sector have the market power
to earn excessive profit margins and drive a price wedge between farmers and consumers. The
analysis is based on data collected along the main matooke trading route from the Western
region to the urban markets around Kampala. This includes quantitative data on the costs,
margins and profits in the different nodes of the value chain; and focus group discussions and
semi-structured interviews with the main actors to understand the trading relationships.
We find no evidence of excessive market power at the intermediary level. Although
intermediaries, particularly wholesalers, do charge relatively high mark-ups, these largely reflect
the high costs of moving matooke from the farmer to the urban consumers. In fact, the markets
between the farmer and the retailer, despite being informal and organic, seem fairly efficient.
Small bicycle collectors pick up the produce at the farm gate and transport it to the district
markets. More remote farmers reach these markets by selling to collectors with motorbikes or
small trucks. Brokers aggregate the matooke at the district markets, from where the produce is
transported by wholesalers to the many small retailers in the urban consumer markets. At every
node there seem to be a sufficient number of alternative buyers to ensure a relatively fair price.
Intermediaries enter the market during the peak season and exit again during off-peak
season. This might explain why profits for intermediaries are kept in check in the peak
season, despite abundant supply. Following this seasonality, our data suggests, however,
there are small shifts in market power: in peak season the intermediaries seem in a better
position, while in off-peak season the farmers seem to exert (some) market power.
More research on competition and market power is, however, required to test the external
validity of these results. Our investigation concentrated on an important and relatively
homogenous staple crop along its main trade route. More remote areas or districts where
there is less of an abundance of matooke might still be best characterised by local
monopsonies where intermediaries have more market power due to high search and transport
costs. Similarly, (local) monopsonies might exist for products for which there is a smaller
market (segment), for products with a stronger seasonal variation in supply or for more
perishable products (see, e.g. Kopp and Mishra, 2022). Moreover, many value chains are not
characterised by the type of spot-market exchange which is prevalent in the matooke value
chain. Instead, value chains are often rather characterised by alternative forms of value chain
governance, involving vertical and horizontal coordination, which, in turn, impacts market
power and distribution of value (Bonanno et al., 2018;Deconinck, 2021;Swinnen, 2020).
The general policy implication of our analysis constitutes a warning: analysing profit
margins and market power is complex and requires taking into account a variety of factors,
such as underlying cost structures and seasonality. Broad-brush statements on intermediary
market power based on observed purchasing and selling prices cannot take these factors into
account. Interventions to improve the trade terms of farmers should therefore be grounded in
a solid analysis of the dominant market structure and prevailing profit margins in the chain.
If not, these interventions might not be relevant to improve trade terms of farmers or, in the
worst case, lead to disruptions and larger inefficiencies.
This likely also applies to the matooke sector in Uganda. Instead of curtailing intermediary
market power, more promising avenues to reduce the matooke price-spread between farmers
and consumers could be explored, including ways to reduce perishability, flattening supply
seasonality and reducing transportation costs. This can, for example, be achieved by upgrading
transportation infrastructure and stimulating the adoption of storage technologies and weather-
resilient agricultural practices.
Market power
in the matooke
value chain
Notes
1. More generally, there is relatively little research taking place on (1) the middle segment of agri-food
value chain (Reardon, 2015), and (2) on the nexus between development economics and industrial
organization (Bellemare, 2022).
2. Mark-ups are commonly defined as the difference between selling price and direct costs (often
expressed in percentage terms). Our computation of mark-up is in line with this definition but we
only include the purchasing price as direct costs.
3. See Akimenyi et al. (2018),Eronmwon et al. (2014), and Giroh et al. (2010) for applications of the
structure-conduct-performance method in African agri-food sectors
4. The total sample size was determined by budget availability.
5. This is September for the collectors and wholesalers and June–July for retailers.
6. Food Balance Sheet from 2018 retrieved from FAOstat on 16 November 2021.
7. Another potential mechanism is related to recency bias. October is typically a month with higher
matooke prices than in July (see Figure 1). Recency bias among our respondents—favouring recent
events over historic ones—can thus potentially have led to underestimation of average prices in
October and overestimation of average prices in July.
References
Abebe, G.K., Bijman, J. and Royer, A. (2016), “Are middlemen facilitators or barriers to improve
smallholders’welfare in rural economies? Empirical evidence from Ethiopia”,Journal of Rural
Studies, Vol. 43, pp. 203-213.
Akinyemi, S.O.S., Adejoro, M.A., Layade, A.A. and Adegbite, O.O. (2018), “Market structure and
performance for plantain and banana in the tropic”,Acta Horticulturae, Vol. 1225, pp. 359-366,
doi: 10.17660/ActaHortic.2018.1225.51.
Ariho, A., Jeremia, M., George, T. and Anna, S. (2015), “Assessment of innovative market access
options for banana value chain in Uganda”,Journal of Development and Agricultural Economics,
Vol. 7 No. 10, pp. 323-331.
Barrett, C.B., Reardon, T., Swinnen, J. and Zilberman, D. (2020), “Agri-food value chain revolutions
in low-and middle-income countries”,Journal of Economic Literature, Vol. 58, pp. 1-67.
Bellemare, M.F. (2022), “Agricultural value chains: towards a marriage of development economics and
industrial organisation?”,Australian Journal of Agricultural and Resource Economics, Vol. 66
No. 2, pp. 241-255.
Bonanno, A., Russo, C. and Menapace, L. (2018), “Market power and bargaining in agrifood markets: a
review of emerging topics and tools”,Agribusiness, Vol. 34 No. 1, pp. 6-23.
Chau, N.H., Goto, H. and Kanbur, R. (2016), “Middlemen, fair traders, and poverty”,The Journal of
Economic Inequality, Vol. 14 No. 1, pp. 81-108.
Deconinck, K. (2021), Concentration and market power in the food chain, OECD Food, Agriculture and
Fisheries Papers, No. 151, OECD Publishing, Paris, doi: 10.1787/3151e4ca-en.
Eronmwon, I., Alufohai, G.O. and Ada-Okungbowa, C.I. (2014), “Structure, conduct and performance of
plantain marketing in Edo State, Nigeria”,Journal of Applied Sciences and Environmental
Management, Vol. 18 No. 3, pp. 437-440.
Fafchamps, M. and Gabre-Madhin, E.Z. (2001), Agricultural markets in benin and malawi: the operation
and performance of traders, Policy Research Working Paper No. 2734, World Bank,
Washington, DC, available at: https://openknowledge.worldbank.org/handle/10986/19415
Fafchamps, M., Gabre-Madhin, E.Z. and Minten, B. (2005), “Increasing returns and market efficiency
in agricultural trade”,Journal of Development Economics, Vol. 78 No. 2, pp. 406-442.
Gabre-Madhin, E.Z. (2001), “The role of intermediaries in enhancing market efficiency in the Ethiopian
grain market”,Agricultural Economics, Vol. 25 Nos 2-3, pp. 311-320.
JADEE
Giroh, D.Y., Umar, H.Y. and Yakub, W. (2010), “Structure, conduct and performance of farm gate
marketing of natural rubber in Edo and Delta States, Nigeria”,African Journal of Agricultural
Research, Vol. 5 No. 14, pp. 1780-1783.
Haggblade, S. and Dewina, R. (2010), “Staple food prices in Uganda”,Paper presented at the Comesa
policy seminar on “Variation in staple food prices: Causes, consequence, and policy options”,
25-26 January 2010, Maputo, Mozambique, available at: https://core.ac.uk/download/pdf/
6449502.pdf
Kalule, S.W. and Kyanjo, J.L. (2013), “Marketing margins and efficiency of cooking banana retail trade
in Kampala City, Uganda”,International Journal of Sales and Marketing, Vol. 3 No. 4, pp. 9-18.
Kikulwe, E.M., Okurut, S., Ajambo, S., Nowakunda, K., Stoian, D. and Naziri, D. (2018), “Postharvest
losses and their determinants: a challenge to creating a sustainable cooking banana value chain
in Uganda”,Sustainability, Vol. 10 No. 7, p. 2381.
Komarek, A. (2010), “The determinants of banana market commercialisation in Western Uganda”,
African Journal of Agricultural Research, Vol. 5 No. 9, pp. 775-784.
Kopp, T. and Mishra, A.K. (2022), “Perishability and market power in Nepalese food crop production”,
Journal of Agricultural Economics, Vol. 73 No. 2, pp. 518-540.
Kopp, T. and Sexton, R.J. (2021), “Farmers, traders, and processors: buyer market power and double
marginalization in Indonesia”,American Journal of Agricultural Economics, Vol. 103 No. 2,
pp. 543-568.
Minten, B., Assefa, T. and Hirvonen, K. (2017), “Can agricultural traders be trusted? Evidence from
coffee in Ethiopia”,World Development, Vol. 90, pp. 77-88.
Mitra, S., Mookherjee, D., Torero, M. and Visaria, S. (2018), “Asymmetric information and middleman
margins: an experiment with Indian potato farmers”,Review of Economics and Statistics,
Vol. 100 No. 1, pp. 1-13.
Nalunga, A., Kikulwe, E., Nowakunda, K., Ajambo, S. and Naziri, D. (2015), “Structure of the
cooking banana value chain in Uganda and opportunities for value addition and postharvest
losses reduction”, RTB-ENDURE Technical report, available at: http://www.rtb.cgiar.org/
endure
Rahman, M.C., Pede, V., Balie, J., Pabuayon, I.M., Yorobe, J.M. and Mohanty, S. (2020), “Assessing the
market power of millers and wholesalers in the Bangladesh rice sector”,Journal of Agribusiness
in Developing and Emerging Economies, Vol. 11 No. 3, pp. 280-295.
Reardon, T. (2015), “The hidden middle: the quiet revolution in the midstream of agrifood value chains
in developing countries”,Oxford Review of Economic Policy, Vol. 31 No. 1, pp. 45-63.
Rietveld, A.M., Ajambo, S. and Kikulwe, E. (2016), “Economic gain and other losses? Gender relations
and matooke production in Western Uganda”,Paper presented at Tropentag 2016: Solidarity in
a competing world —fair use of resources, September 18-21, Vienna, Austria.
Steijn, C., Kruijssen, F., Smits, E., Mulumba, N., Asindu, M., Kuijpers, R. and Kikulwe, E. (2022)
“Quantity and quality of employment in Uganda’s matooke value chain”, Working paper (not
submitted).
Swinnen, J. (2020), Competition, market power, surplus creation and rent distribution in agri-food value
chains, Background paper for The State of Agricultural Commodity Markets (SOCO) 2020,
FAO, Rome.
Uganda Bureau of Statistics (2020a), Uganda Annual Agricultural Survey 2018, UBOS, Kampala,
Uganda, available at: https://www.ubos.org/wp-content/uploads/publications/AAS_2018_
Report_Final_050620.pdf
Uganda Bureau of Statistics (2020b), “Agricultural production of Uganda [total production of major
crops - plantain bananas]”, available at: https://uganda.opendataforafrica.org/kdcloe/
agricultural-production-of-uganda (accessed 20 November 2020).
Market power
in the matooke
value chain
Warsanga, W.B. and Kilimanjaro, M. (2014), “Coordination and structure of agri-food value chains:
analysis of banana value chain strands in Tanzania”,Journal of Economics and Sustainable
Development, Vol. 5 No. 7, pp. 71-78.
Xhoxhi, O., Stefanllari, A., Skreli, E. and Imami, D. (2020), “How intermediaries’power affects contract
farming decisions: evidence from the medicinal and aromatic plant sector in Albania”,Journal
of Agribusiness in Developing and Emerging Economies, Vol. 10 No. 5, pp. 529-544.
Yi, J., Meemken, E.M., Mazariegos-Anastassiou, V., Liu, J., Kim, E., G
omez, M.I., Canning, P. and
Barrett, C.B. (2021), “Post-farmgate food value chains make up most of consumer food
expenditures globally”,Nature Food, Vol. 2 No. 6, pp. 417-425.
Zijlstra, S. (2015), “Uganda’s cash machine”,Vice Versa, Vol. 48, pp. 26-31.
Corresponding author
Rob Kuijpers can be contacted at: r.kuijpers@kit.nl
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com
JADEE