The hidden middle: The quiet revolution in the midstream of agrifood value chains in developing countries

Abstract

The food security debate has focused largely on the farm sector and on trade. Relatively neglected or ‘hidden’ from mainstream
debate are the middle segments (processing, logistics, wholesale) of agrifood value chains in developing countries—and yet
this ‘midstream’ forms 30–40 per cent of the value added and costs in food value chains. The productivity of the midstream
is as important as farm yields for food security in poor countries. The article shows that over the past several decades the
middle segments have transformed quickly and surprisingly—with a huge volume expansion, a proliferation of small and medium
enterprises (SMEs), but also concentrating and multinationalizing (in some places and products), with technology change characterized
by capital-led intensification, and with the incipient emergence of branding and labelling and packaging, of new organizational
arrangements in procurement and marketing interfaces with farmers and retailers, and of private standards and contracts. Economic
policies of market and foreign direct investment (FDI) liberalization, commercial and business climate regulations, hard and
soft infrastructure investment, and food safety laws, have paved the path to the expansion and shaped the transformation of
the important midstream segment of food value chains.

Full text

doi:10.1093/oxrep/grv011
© The Author 2015. Published by Oxford University Press.
For permissions please e-mail: journals.permissions@oup.com
The hidden middle: the quiet revolution in
the midstream of agrifood value chains in
developing countries
ThomasReardon*
Abstract The food security debate has focused largely on the farm sector and on trade. Relatively
neglected or ‘hidden’ from mainstream debate are the middle segments (processing, logistics, whole-
sale) of agrifood value chains in developing countries—and yet this ‘midstream’ forms 30–40 per cent
of the value added and costs in food value chains. The productivity of the midstream is as important
as farm yields for food security in poor countries. The article shows that over the past several decades
the middle segments have transformed quickly and surprisingly—with a huge volume expansion, a
proliferation of small and medium enterprises (SMEs), but also concentrating and multinationalizing
(in some places and products), with technology change characterized by capital-led intensication, and
with the incipient emergence of branding and labelling and packaging, of new organizational arrange-
ments in procurement and marketing interfaces with farmers and retailers, and of private standards
and contracts. Economic policies of market and foreign direct investment (FDI) liberalization, com-
mercial and business climate regulations, hard and soft infrastructure investment, and food safety laws,
have paved the path to the expansion and shaped the transformation of the important midstream seg-
ment of food value chains.
Key words: value chains, developing countries, Asia, wholesale, wholesale markets, food processing,
structural change, food industry transformation
JEL classication: O12, O13, O16, O19, Q12, Q13, Q18
I. Introduction
Academic debates on changing food markets in developing countries have tended to
focus on three main themes over the past several decades. The rst theme is the growth
of exports and imports with trade liberalization and globalization (e.g. Anderson etal.,
1997), and sourcing by US and European supermarkets from developing countries
(Dolan and Humphrey, 2000). The second theme is the transformation of upstream
in the food system, in farming intensication, commercialization, and diversication
(Pingali and Rosegrant, 1995) and the growth of input markets, for water (Rosegrant
* Michigan State University, e-mail: reardon@msu.edu
The author gratefully acknowledges nancial support from the World Bank and USAID through its
Food Security Policy Innovation Lab, and useful comments from Christopher Adam and Doug Gollin.
Oxford Review of Economic Policy, Volume 31, Number 1, 2015, pp. 45–63

ThomasReardon
46
et al., 1995), land (Deininger and Feder, 2001), and improved seeds (e.g. Pray and
Naseem, 2007). The third theme is transformation downstream in the domestic food
system, especially the ‘supermarket revolution’ (Reardon etal., 2003; Traill, 2006) and
diet diversication and shift towards highly processed foods with attendant health chal-
lenges (Pingali, 2006; Popkin, 2014).
In comparison, there has been far less attention paid, in research and more notice-
ably in policy debates, to the rapid transformation of the mid-stream segments of agri-
food value chains,1 by which we mean processing, storage, wholesaling, and logistics.
These segments are increasingly important: for example, Reardon et al. (2012) show
that share of the midstream segments in total margins in rice and potato food value
chains to the capital cities of Bangladesh, China, and India averages 32 per cent for rice
and 42 per cent for potatoes. Those pieces of the food system are too important to stay
‘hidden’ from the debate.
Moreover, as we further explore in this paper, there is emerging evidence that there
is rapid change in the midstream’s structure and conduct/behaviour; we illustrate this
mainly for developing Asia (with comparative mention of similar trends in Latin
America and, in incipience, in Africa). This change is driven mainly by private-sec-
tor investment by both large and small enterprises, encouraged by transformation in
upstream and downstream in the food system, and facilitated by policy reforms. In
an interesting parallel, a similar phenomenon was observed in the mid-1990s in the
US by Schertz and Daft (1994) in their book about the ‘quiet revolution’ in food and
agricultural markets, where they noted how rapid transformation of the midstream seg-
ments had not only occurred rapidly through the 1970s and 1980s, but that it had also
remained hidden to policy debate that in the US had also been focused upstream on the
farm and input sectors.
This paper focuses on the transformation of the midstream of food value chains in
developing countries. We focus on domestic food system transformation rather than
international food value chains involving developing countries, because domestic food
production constitutes roughly 90–95 per cent of the food economy of the region, and
international trade in food is only about 5–10 per cent. We also illustrate mainly with
Asian evidence of midstream transformation, simply because the midstream transfor-
mation started earlier and has gone furthest in Asia and Latin America, with Africa just
starting on what appears to be a similar road of food system transformation (Reardon
etal., 2013); moreover, between Asia and Latin America, there has been more recent
empirical survey work on the midstream transformation inAsia.
The paper is organized into three parts. The rst part sets the stage by discussing
the conditioners of the transformation. The second part focuses on the rst of two
broad axes of the transformation, namely the structural transformation of the mid-
stream which is itself integrated into change in the overall structure of the food sys-
tem (including geographic lengthening combined with dis-intermediation in food value
1 We use the term ‘value chain’ rather than ‘supply chain’ for simplicity of terms; the former has been
associated with quality differentiation and value added, with viewing the value chain from the consumer
perspective, and the latter, supply chains, from the supplier perspective with a focus on efciency and logistics
and coordination aspects of moving products from ‘farm to fork’. However, we believe, along with Feller
etal. (2006), that there should be an integration of these terms and concepts as food systems need to and do
(with varying performance levels) deliver both value and efciency.

47
The midstream of agrifood value chains in developing countries
chains). The nal part focuses on the second of the broad axes of transformation, that
of conduct/behaviour of the midstream segment, including rms’ choice of technolo-
gies, institutions (such as standards and contracts), and organization (such as vertical
and horizontal integration and coordination).
Two caveats are relevant here. First, it is beyond the scope of the paper to assess
how these transformations affect consumers and farmers and other welfare and per-
formance aspects; those are important topics for further research. Second, the paper is
mainly qualitative and illustrative, with the use of as much empirical data as are avail-
able for the patterns; but developing countries in general, with few exceptions, mainly
track crop output and exports but have much less data and detail on the other segments
of the value chain. Data sources on part of the midstream, in particular processing,
like the United Nations Industrial Development Organization (UNIDO), do not offer
sufcient disaggregation to track the midstream segments in detail. So we had to rely
on key indicators of change and case information.
II. Broad determinants of transformation of the midstream
of the valuechains
Reardon and Timmer (2014) develop a framework to describe the transformation of
food systems in terms of ve simultaneous and inter-linked spatio-economic transfor-
mations in Asia: (i) urbanization; (ii) diet change; (iii) transformation of the midstream
and downstream components of the agrifood system; (iv) technological and commer-
cial transformation of the farm segment; (v) transformation of the upstream compo-
nents of the agrifood system with accompanying rural factor market (labour, credit,
land) transformation. They note that urbanization and diet change are the demand-side
forces pulling the whole set of transformations; factor market and farm technology
change are the upstream supply-side forces feeding the rest of the changes; and the
agrifood system spans all the segments and intermediates or links supply and demand.
Because the present paper focuses on the midstream transformation ‘nestled’ among
and conditioned by the upstream and downstream transformations, we briey discuss
here the latter two; but for succinctness and because the downstream forces are more
important than upstream changes in affecting the midstream segments, we start with
a review of recent ndings on urbanization and diet change (and do not treat here the
upstream changes already mentioned in the introduction). Moreover, we outline sev-
eral ‘policy meta conditioners’ of all ve transformations, in particular, public-sector
investments in hard and soft infrastructure, and policy interventions including direct
public-sector intervention in the midstream, and, after intervention, liberalization and
privatization of midstream. We consider those four sets of conditioners inturn.
(i) Urbanization
First, rapid urbanization has emerged quickly and relatively recently in much of Asia.
By 2010 the urban population share reached 32 per cent in south Asia, 44 per cent in
south-east Asia, and 54 per cent in east Asia. But population shares alone underestimate

ThomasReardon
48
the importance of urbanization for the food economy: urban consumers have lower
shares of food in total household expenditure compared with the rural population, but
their incomes are sufciently higher that their per capita food expenditure is higher.
India exemplies this: Ablett et al. (2007) note that by 2006, while only 29 per cent
of India’s population was in cities, urban consumers accounted for 43 per cent of all
market expenditures on food consumption, while in more urbanized countries of east
and south-east Asia, urban consumers are responsible for roughly two-thirds to even
three-quarters of all food expenditures. For example, 74 per cent of fruit in Indonesia is
consumed in cities (Reardon and Timmer, 2014; Reardon etal., 2014a).
The process of urbanization conditions the rural non-farm economy and thus in
turn the rural part of midstream activities. On the one hand, there appears to be more
rapid growth of the rural non-farm economy in the market catchment areas of mega
and intermediate cities (see, for example, Deichmann etal. (2009) for Bangladesh and
Fafchamps and Shilpi (2003) for Nepal). This ‘radiation’ is magnied as a positive func-
tion of rural infrastructure. The correlate is that in more hinterland areas the process is
occurring less quickly.
On the other hand, much rural non-farm activity is directly or indirectly linked to the
midstream segments, such as transport and wholesale of agrifood products (Haggblade
etal., 2007). But the radiating effects of urbanization can also cause competition for
food-system-related rural non-farm employment. Reardon etal. (2007b) note that this
competition can be especially in the midstream segments, as the intermediate cities are
directly (themselves) and indirectly (as conduits for products from mega-cities) a chan-
nel for urban manufactured food products and services into rural areas, pitting urban
rms with economies of scale against small traditional food rms in the informal urban
and ruralareas.
Moreover, recent research (such as Reardon etal. (2012) in Asia) tends to show that
areas nearer cities experience more rapid transformation of food value chains, including
the development of the midstream. Acity’s food market catchment area can be national
or further, but it appears that in a certain radius the city’s demand pull is strongly felt and
the prots from the urban market induce substantial local investment by midstream rms.
(ii) Diet transformation
Second, diets have been rapidly changing in Asia (Pingali, 2007)2 in three ways, all
important for the development of the midstream segments.
The rst change in diets, as incomes have risen, has been a shift in the food expendi-
ture shares, with a decline in the share of rice (Timmer and Dawe, 2010) and an increase
in the shares of meat/sh, dairy, horticulture products, and edible oils, and an increase in
corn and soya use for feeding animals. This is consistent with ‘Bennett’s Law’ (Bennett,
1954). In India, for example, by 2010 the total value of dairy consumption was greater
than that of grains (Kumar etal., 2013). This has happened earlier and faster in urban
areas, inducing rapid development of wholesale and logistics from rural to urban areas
for products intensive in handling.
2 Diets have been changing and urbanization proceeding in a similar way, just with a lag, in Africa, see
Tschirley etal. (2015) for eastern and southern Africa and Hollinger and Staatz (2014) for West Africa.

49
The midstream of agrifood value chains in developing countries
The second change in diets is that rural consumption has rapidly ‘commercialized’ (i.e.
the share of purchased food in total food consumed has risen). This bolsters demand
for marketing and logistic services, again part of the midstream. Mellor (1976) noted
that many rural households beneted from a reduction in food prices due to the Green
Revolution in India because so many were net buyers. This left unanswered how impor-
tant purchases are in the food budget, and thus how important rural marketing is from
the demand side. New evidence shows that it is very important. Reardon etal. (2014a),
based on analysis of Living Standards Measurement Study (LSMS) data in Bangladesh,
Indonesia, Nepal, and Vietnam, show that rural households on average buy a large share
of their diet. Purchased food in total food consumed was found to be about 80 per cent in
rural Bangladesh and Indonesia, 72 per cent in rural Vietnam, and 58 per cent inNepal.
The trend counterpart to an increase in purchased food by rural households is the
rise in the generation of cash by rural households, from rural non-farm employment
as noted above, and from commercialization of farming (Pingali and Rosegrant, 1995).
The latter has proceeded very far; Reardon etal. (2014b) found in surveys of rice farm-
ers in Bangladesh, China, India, and Vietnam, that the marketed surplus rate is about
85 per cent on average, even very high for very smallfarms.
The third change in diets is that, with the rise of incomes and the rising opportunity
cost of women’s time with urbanization, there has been a rapid penetration of both
low and high processed foods (such as wheat noodles, see Timmer (2015)), as well as
prepared foods sold in restaurants. This has occurred rst and most forcefully in urban
areas but with a lag and only somewhat less in rural areas. This is a major llip for the
supply-side development of processing and marketing as part of the midstream. For
example, Reardon et al. (2014a) nd that even in rural areas processed food has dif-
fused, with 59 per cent of total rural food expenditure in Bangladesh, Indonesia, Nepal,
and Vietnam accounted for by processed food, of which 69 per cent is ‘low processed’
and 31 per cent ‘high processed’. The overall processing share rises with development
level and with the degree of urbanization: the simple average over the four countries for
urban areas was 73 per cent of total food expenditures. Morisset and Kumar (2008) had
a similar nding for India.
(iii) Infrastructure investment
Third, government investment in hard and soft rural infrastructure has helped to
increase the length and volume and reduce the seasonality of food value chains. In
turn, this has encouraged urban diet change and magnied impacts on rural areas.
‘Distance’ to urban areas is not just a physical distance but an economic distance, in
the sense that road infrastructure, toll highways, and train and bus routes condition the
transport cost of a given distance. In most countries in the region there have been large
expenditures by governments on transport and electricity infrastructure in the past
several decades.3 Wholesale market infrastructure is another important public-sector
investment. For example, China’s wholesale market volume increased 11,000 per cent
from 1990 to 2000 (Huang etal., 2007; Ahmadi-Esfahani and Locke, 1998), and India’s
3 See for example Fan and Chan-Kang (2005) for analysis of massive rural road investments in China
in the 1980s–1990s.

ThomasReardon
50
regulated wholesale markets went from 450 in 1948 to 5,500 in 2008 (Reardon et al.,
2012). The latter has been reinforced by the establishment of the ‘soft infrastructure’ of
commercial regulations and public standards.
There have also been a large number of modest investments from the small–medium-
scale private sector—by truckers, warehouse owners, millers, cold storage operators,
wholesale traders, and rural brokers. These have been mainly aimed at accessing the
urban market; even in Vietnam, a country where rice exports are substantial, most of
the investments and transformation are occurring in the domestic rice value chain, not
just the minor part of the rice sector which is traded.
(iv) Government policy: first intervention then liberalization
Fourth, there have been direct and indirect government market interventions in the
agrifood system, including in the midstream sector. Arst ‘round’ included public- and
large private-sector investments before the 1950s from colonial enterprises, especially in
south-east Asia (for the export-oriented agrifood industrial enclaves and plantations),
and in a new wave in the 1950s–1970s by national governments. Upstream, these inter-
ventions included government input and credit parastatals; at the midstream with the
provision of public wholesale markets (rst for grains and then for ‘wet goods’ such as
sh, meat, and fruits/vegetables) and distribution agencies and processing parastatals;
and downstream with export marketing boards; and some public-sector retail facilities
such as the state grain stores in India andChina.
A ‘second round’ of food system transformation occurred following the withdrawal
by most governments in the region from direct intervention. While the privatization of
agricultural parastatals and trade liberalization have received the most public attention
(see Rashid etal., 2008), an equally important part of this round was the opening up of
the sector to FDI, including substantial ows into the midstream segment. Asia has thus
seen cross-border investments from large global rms such as Cargill, and regional rms
such as Thailand’s Charoen Pokphand (CP), the Philippines’ San Miguel, and China’s
Shuanghui Group (that in 2013 bought what was the largest pork processor in the world,
Smitheld Foods of the US), which have paved the way for a surge in new investments
as well as merger and acquisition activity by large domestic rms and eventually SMEs.
III. Transformation in the midstream
In this section we explore the dynamics of the transformation of the midstream agri-
food sector in developing Asia from ‘traditional’ to ‘modern’ by examining change in
the structure and conduct of the overall food system and of the midstream segment in
particular.
A qualication is in order before we lay out the axes of transformation. The trans-
formation itself has exhibited a substantial degree of variation over space and products
in the speed and form of this transformation depending in part on countries’ overall
levels of development. The transformation has taken place unevenly over countries in
the Asian region (as in other developing regions). In the earlier phase, from the 1950s
to the 1980s, this unevenness was less acute in the rst round (1950s–1980s) when the

51
The midstream of agrifood value chains in developing countries
primary drivers were public-sector direct action (such as the erection of parastatals).
By contrast, since the 1980s when the key drivers were urbanization and liberalization,
experiences across countries were less evenly spread.
The process of the structural transformation of midstream value chains roughly
reects the order/ranking of countries in terms of urbanization, industrialization, lev-
els of income per capita, and liberalization of policy. Thus the process started in Japan
and Taiwan and South Korea followed by the early-liberalizing countries of south-
east Asia, Malaysia, Indonesia, the Philippines, and Thailand. These countries were
followed by the transition-cum-late-liberalizing countries of China and Vietnam and
nally India and the smaller poorer countries of Cambodia, Laos, and Myanmar.
There is also unevenness in the transformation within countries, particularly between
rural and urban sectors. But, as noted above, within rural areas there is a sharp dif-
ference in the progress of the transformation processes in dynamic/near-city zones
compared with remote, hinterland areas. The exception is development of rst-stage
processing enclaves in the hinterland areas (such as for tea, rubber, oil palm, and cassava
processing in large plantations in frontier areas of Cambodia, Laos, and Myanmar; see
Byerlee, 2014).
Moreover, there are many instances of a ‘migration’ of the transformation into the
hinterland areas as the dynamic urban areas ‘climb the value ladder’ and the produc-
tion of cheap food products and rural processing and wholesale are pushed further out
to areas with cheaperland.
The transformation of both the retail segment and the midstream segment occurs
in general at different rates over product categories. The rst wave was in grains and
traditional export crops, where transformation was driven by government-led reforms,
and in plantation cropping and processing, in tea, rubber, and oil palm. Next came
processed foods in general, and grains and edible oils and condiments such as soy sauce
in particular—products that could be shipped and stored at low cost—followed by
the feedstock sector linked to the booming demand for livestock, sh and, eventually,
milk. Next were semi-processed perishables such as meat, sh, and dairy products, and
nally, fresh fruits and vegetables. The economic logic of the sequence is a combination
of perishability (and thus the challenges that brings to value chains) and the emergence
and growth of substantial demand in urban areas for these products in direct or derived
demand.
Although the above discussion emphasizes the sheer heterogeneity of the transfor-
mation, there is nonetheless a series of clear patterns of change in developing Asia (mir-
rored in other developing regions) in the structure and conduct of food value chains in
general and the midstream as our particular focus. These are discussedbelow.
(i) Structure change: spatial elongation and de-seasonalization of
valuechains
We noted above how urbanization combined with rural infrastructure improvement
(and very low dependence on imports) make it that food value chains need to get longer
and longer within a country, or for smaller countries within a sub-region, to supply
the growing large cities and the rapidly emerging intermediate cities. This means that
the logistics/transport and rural and urban wholesale sub-segments expand in parallel

ThomasReardon
52
with the spatial lengthening of the value chain. The China rice sector is illustrative:
the share of the midstream in the value chain of rice to Hangzhou in southern China
mainly from southern and middle regions is about 25 per cent of total margins (Wang
etal., 2013); for the much longer value chain to Beijing from the north-east provinces
the share of the midstream is about 50 per cent (Reardon etal., 2010).
Von Thünen (1826) founded our now-traditional conceptualization of the standard
pattern of correlations between spatial length of a rural–urban value chain and prod-
uct categories. He noted that farm product composition changed over space as one
went away from a city, as a function of land and other farm costs and transport costs;
von Thünen noted that perishable, high-value products, such as vegetables and dairy,
tended in that era to be produced in a concentric ring near cities, followed by a ring for
timber, then extensive grains, and then hinterland.
While these spatial-cum-product composition patterns were also typical in develop-
ing countries as recently as several decades ago, and in the poorest areas are still, they
are beginning to change, partly as a function of the transformation of the midstream
segments. The latter has begun to alter the geography and seasonality of farming and
value chains to cities. We briey discuss this by midstream sub-segment as follows.
(1) Development of transport via road and rail has extended the market catch-
ment area of rice and wheat, and of semi-perishables such as potatoes and
apples.
(2) Development of the sh freezing industry has recently spurred long domestic
value chains in sh, for example in China from southern China to Beijing.
(3) Development of the poultry, hog, and sh feed industry has allowed intensive
production further or even far from cities; for example, Yi etal. (2015) discuss
the rise of intensive shrimp farming off Java, in islands with improved water
conditions; feed is produced on Java, sent by boat to these islands, then shrimp
is harvested and shipped to export points and cities. Fish farming is blossoming
in clusters around Bangladesh, with feed produced near Dhaka then trucked
to these clusters and sh put in barrels of ice and shipped back to Dhaka.
NaRanong (2008) tells a similar story for chicken production in Thailand.
(4) Development of medium and large-scale output processing has allowed both
economies of scale and location of plants further from cities; Schneider (2011)
notes for China that as small-scale pig production dropped from 95 to 27 per
cent of pigs over 1985 to 2007, the scale of pig processing plants soared and
moved away from peri-urban areas. Highly processed foods made by Indofoods
in Jakarta or San Miguel in Manila are consistently available in most parts—at
least of the main islands, where three-quarters of the food economies are—of
Indonesia and the Philippines, respectively.
(5) Development of medium and large-scale feed processing mills in maize and
soya areas has induced the movement of chicken production from peri-urban
areas out to these crop areas, such as in Thailand in the 1990s (Rushton etal.,
2005).
(6) Development of the cold storage industry has recently and quickly expanded
and de-seasonalized value chains of potatoes, for example, to cities in India
(Das Gupta etal., 2010) and Bangladesh.

53
The midstream of agrifood value chains in developing countries
(ii) Structure and conduct change: at first a proliferation of
intermediaries, followed by dis-intermediation in valuechains
In Asia, as in other developing regions, the traditional foundations of the food value
chain were rooted in farmers’ sales to a nearby village or rural town or neighbour.
As rural towns and cities grew, there was a proliferation of intermediaries in the mid-
stream; rural–urban brokers emerged in dendritic structures to collect from rural areas
then sell on to villages then to semi-wholesalers in district towns who then on-sold to
cities (see, for example, Skinner (1964) for China or Lele (1971) for India).
The subsequent stage, starting in the 1960s/1970s, has been a ‘rst wave of dis-inter-
mediation’ in the midstream. As rural–urban value chains developed and lengthened,
and intermediaries proliferated, the image grew in the policy debate and research lit-
erature in the Asian region (and other developing regions) of ‘many hands’ stretching
along the long rural–urban value chains. Often in policy debates the terms ‘speculation,’
‘exploitation’, and ‘inefciency’ became attached to rural brokers and wholesalers; that
image (empirically supported or not) served to spur support for the establishment of
marketing and processing parastatals in the 1960s–1970s as a form of ‘state-led dis-
intermediation’ of grain value chains to cities.
The emerging evidence points to a process in the 1980s–2000s in which state establish-
ment of parastatal processors and marketing in the 1960s–1970s and investment in rural
and urban wholesale markets and roads in the 1970s–1990s in urban and then rural
areas, led to a ‘second wave’ of grass-roots private-sector-led restructuring of rural dis-
tribution. Wholesalers from rural wholesale markets and even urban markets began to
buy directly (via transporters) from farmers; mills got bigger and started to buy directly
from farmers; trucking rms proliferated and eased local logistics constraints. The result
in many places has been that small rural brokers have been progressively sidelined then
eliminated, with many recent surveys showing this in rice, mangoes, tomatoes, potatoes,
and other; see for example Huang etal. (2007) for tomatoes in Shandong.
A ‘third wave’ of disintermediation in value chains and restructuring of the distribu-
tion segment has emerged in the 1990s and especially the 2000s. This can be seen from
two angles, procurement and marketing.
From the angle of procurement, the third wave has involved the emergence—either
from inside wholesale markets or outside wholesale markets—of modern ‘specialized-
dedicated wholesalers’ (Reardon and Berdegue, 2002). These buy direct from rst- and
second-stage processors and farmers as agents of supermarkets (such as Bimandiri on
Java, see Natawidjaja etal. (2007)) and processors. On the other hand, this has involved
midstream and downstream rms buying directly from each other (such as large retail
chains in Beijing buying directly from large mills in north-east China). This latter is
sometimes facilitated by the rms of two segments of the food industry delivering
between their ‘distribution centres’ to reduce transaction costs. This can extend over
countries in the region to affect international trade by ‘internalizing’ trade in intra-rm
transactions as in ‘new trade theory’ (see Reardon etal., 2007a).
Dis-intermediation of local procurement agents can be supplanted by—reinterme-
diated—via enlisting the assistance of non-local agents. An emerging and important
form of this is ‘follow sourcing’, where product or service providers for a food industry
rm ‘follow’ that rm to the new market to which that rm locates. An example is
TNT Logistics ‘following’ Tesco to Thailand, or Baakavor ‘following’ Tesco to China

ThomasReardon
54
4 In recent Asian literature the example of vegetables presents itself; cf Gorton etal. (2011) for Thailand;
Moustier (2009) for Vietnam; as does that of chicken and pork (with Thailand having gone very far in the
concentration from the late 1970s to present, see NaRanong (2008); China being in mid-path (see Schneider,
2011), and, with the case of dairy, India being at the early stage with small and marginal farms still producing
68 per cent of the milk but medium/large private-sector rms undertaking 58 per cent of the processing of
marketed milk by 2010 (Kumar etal., 2013).
(Reardon et al., 2003, 2007a). This ‘following’ also occurs across provinces, as food
industry rms spread their operations over zones; an example is cold chain services sup-
pliers in China following KFC diffusion in China (Bell and Shelman, 2010).
From the angle of marketing, the third wave of dis-intermediation has involved large
processors’ obviating the traditional channel of broker-stockists to distribute directly
to small retail stores (as well as to supermarkets) with their own delivery vans or third-
party logistic companies. An example is Yili distributing its dairy products directly to
shops and supermarkets in Beijing (Abrami etal., 2008).
Where the product is perishable this appears to be increasingly done with the help of
integrated logistics rms undertaking a variety of tasks—wholesaling (intermediation),
warehouse management, ICT system integration into retail and distribution systems of
companies, cold chain development, and packaging. They may also forward-integrate
into retail management of specic divisions (such as Radhakrishna Foodland in India
becoming an external ‘channel captain’ managing fresh produce for second-tier India
supermarket chains). This has been via multinational companies such as the Japanese
Snowman Frozen Foods Ltd. It has also occurred via domestic investment which has
been emerging in this sub-segment, sometimes from transport company roots (such
as CONCOR in India for the rail segment), maritime company roots (such as Adani
did in India), and hotel roots—in short, companies that had some transport functions
that then were extended into logistics for modern agrifood companies. Some were con-
glomerates that had food operations and saw the unmet demand for modern logistics
and added logistics; in India, Pantaloon (the leading retailer) started a major logistics
company (see Reardon etal. (2012) for India cases).
(iii) Structural change: aconsolidation of the midstream segments
of the valuechains
Consolidation in the midstream segment of agrifood value chains has followed a num-
ber of paths acrossAsia.
The rst path is a monotonically increasing concentration curve where the segment
starts (traditionally) fragmented, as a set of small-scale rms, and then gradually or
quickly concentrates because of foreign or local investment by large rms or by organic
growth of the small rms—but without passing through a stage of state-induced con-
solidation via parastatal formation. Most perishable product value chains are examples
of this—chicken, sh, pork, and fruit and vegetables.4
The second path, such as one nds in the rice sector, is a J-curve over time of concen-
tration (with time on the horizontal axis and concentration on the vertical axis), where
the traditional phase is as usual only small-scale rms, and the rst stage of consolida-
tion is induced by government establishment of parastatals. However, the parastatals
dominate only a part of the market; alongside them are small-scale operators (informal

55
The midstream of agrifood value chains in developing countries
or formal depending on whether the ‘parallel market’ is legal) (Roemer and Jones,
1991). After the dissolution of the parastatals there was then sometimes a proliferation
of small formal or informal rms in theirstead.
A further (nal?) phase of consolidation in the J-curve (or the U-curve below) is the pri-
vate-sector-led consolidation after liberalization from FDI and from large extra-regional
capital (such as Cargill), large regional capital (such as Thailand’s Charoen Popkhand’s
investments in Cambodia), and large domestic capital. In turn, domestic or regional rms
grown large by this process are recently beginning to make the same kind of jump to
extra-regional investments as did the US and western European companies after a similar
but earlier process of consolidation in their regions. An example is the recent purchase by
China’s pork farming and processing giant, Shuanghui, of what had been the largest pork
processing/production rm in the world, the US-based Smitheld Foods (Xia, 2015).
This latter wave, private-sector led, of consolidation of the midstream features the
pushing out or acquisition or sometimes mergers of large midstream rms with the
small and medium rms. In a sense, the new wave of concentration in the midstream
is an expanded but repeated version of the earlier state-led concentration of the mid-
stream, but now mainly with private-sector dominance. There are exceptions to the lat-
ter, such as in giant state-owned processing rms (such as COFCO in China; Collinson
and Rugman (2007)) and government distribution of grain in India (but extremely little
in other Asian countries) (Rashid etal., 2008).
A third path is a U-curve over time of concentration. This is where a sector has
been traditionally concentrated; then there may be a phase of relative de-concentra-
tion, either from liberalization or from new zones/areas being brought into production;
nally there is a phase where there are mergers and acquisitions from large domestic
capital or multinationals. Illustrations of this are, of course, the estate/export crops in
several countries; rubber in Myanmar is an example (Byerlee etal., 2014). An interest-
ing variation on this third path is where concentration may be occurring in the ‘com-
modity’ branch of the product while in parallel there is a proliferation of small and
medium rms producing differentiated products to compete with the commodity. An
illustration of this is a concentrated brewing sector with a ‘commodity’ beer, parallel to
a number of micro-breweries offering quality-differentiated products.
Note that some parts of the midstream can be concentrating while others are lag-
ging, or the overall midstream is concentrating while other segments of the value chain
lag in concentration or are even fragmenting further. For example, tea plantations and
processing, say, in Indonesia can be concentrated, but at least initially the retail of tea
can be of low concentration, through many small shops. It appears that in most sectors
in developing Asia (and other developing regions), processing concentrates rst (due
mainly to economies of scale in processing), followed by retail, and then by wholesale.
However, the unevenness of concentration over parts of the midstream and over seg-
ments of the value chains appears gradually to be diminishing. This is driven by shared
drivers of concentration, as well as competition in the post-liberalization era after the
1980s. Moreover, there is evidence that the concentration in one value chain segment
‘induces’ concentration in the others through favouring scale in choice of client or sup-
plier. This can be termed ‘symbiosis’ among the modern agrifood industry segments,
as follows.
On the one hand, large processors reduce transaction costs for modern retailers by facil-
itating dis-intermediation, delivering to the distribution centres or stores of the retailers

ThomasReardon
56
(such as large rice mills in north-east China delivering directly to supermarket chains in
Beijing; see Reardon etal., 2012). Large processors can adapt packaging and variety to
the needs of the retailers; their inventory systems reduce the chance of retail stock-out.
On the other hand, modern retailers facilitate development of market size and scope
economies for large processors. Supermarkets tend to carry a limited set of brands per
product category, and these tend to be mainly from medium and large processors, and
a smattering of small company brands for non-commodity products. Modern retailers
develop markets for processed products as they tend to sell them cheaper than tradi-
tional stores once procurement systems are modernized (Minten etal. (2010) for India).
Additionally, large processors and supermarket chains provide the initial key markets
for modern wholesalers (the ‘dedicated wholesalers’ noted above) and modern logis-
tics rms. These rms are competing with traditional wholesalers to serve the modern
retailers and processors—and do so by offering often better transport services (with
modern cross-docking and refrigerated vehicles), warehousing management, and ser-
vices not usually found in traditional distribution segments, such as operating packing
houses, packaging, ICT systems, and cold chains, and managing contract farming, mer-
chandise inventory, and international networks.
(iv) Conduct change: ashift from labour-intensive to capital-
intensive technologies in midstreamfirms
Across many developing Asian countries, there has been a shift from small-scale to
increased scale of processing plants, and of wholesale and logistics and storage operations.
This is roughly correlated with the concentration paths, as well as the waves of transforma-
tion over countries and products noted above. For example, there were large rice process-
ing units during the state-led period in several of the countries; then after privatization
there was a proliferation of smaller plants; and then with technological change (and con-
solidation), there was an exit of smaller mills (such as in rice in China and Vietnam) and
a rapid rise of larger mills owned by large rice milling companies (Reardon etal., 2012).
The increase in scale and the capital/labour ratio has been driven by several factors
beyond the most obvious one that there are economies of scale in processing and com-
petition drives rms to seek those economies.
First, there have been large inows of investment in xed plant and equipment in the
processing sector in the past two decades. For example, investment in xed plant in this
segment in China jumped from 26.3 to 84.7 billion USD from 2007 to 2011 (National
Bureau of Statistics of China, 2012), a ratio of 3.2, while the national nominal GDP rose
in that period by a factor of 2.2. China and India explicitly encourage technology upgrad-
ing and plant size increase in measures for the food processing industry in their 12th
5-year Plans (covering these past 5years). Domestic investment is encouraged by policies
cheapening credit (as it had been in the Indonesian rice sector in the 1970s; see Timmer,
1973) as well as by subsidies for land acquisition and xed period tax exoneration.
That rise in investment was also partly due to the liberalization of foreign investment
in food processing in the 1990s and 2000s. For example, China and India liberalized
food processing FDI and large inows occurred. FDI into food processing in India had
averaged only US$117m per year during 2001–12, then went to US$401m in 2012–13,
and then in the rst half of 2014–15 jumped to US$2.15 billion (Times of India, 2014).

57
The midstream of agrifood value chains in developing countries
Second, it appears that food safety regulations increased plant size and induced mod-
ernization of equipment in developing Asia. This impact is like that of the 1906 Pure
Food and Drug Act in the US on processing and distribution rm scale in the US:
inducing rapid exit of small rms unable to meet the new requirements (Levenstein,
1988). For example, regulations concerning hygiene and location of poultry processing,
production, and retail facilities were put in place during the bird u crisis in Vietnam
and spurred the development of larger-scale and formal plants such as those of Chaoren
Popkhand (Figuié etal., 2013). Health regulations had a similar effect on poultry pro-
cessing in Thailand (McLeod etal., 2009).
Sometimes regulation resisted transformation: for example, India ‘reserved’ most of
food processing for small enterprises until 1998, to protect employment. In 1998, as
part of overall liberalization, the sector was ‘de-reserved’—and a ood of investment
quickly increased the concentration indices and deepened capital (Bhavani etal., 2006).
Moreover, while the modernization and increase in scale of plant and equipment in
the midstream is more ‘obvious’ and observable in the case of large rms, it is important
to note that there appears to be a massive amount of investment by small and medium-
scale rms. This is emphasized in Reardon etal. (2012), with widespread investment by
small transporters, cold storage and warehouse operators, wholesalers, and processors
in the past 10–15years. This corresponds to the discussion above of the ‘proliferation
of SMEs’ phases in structural change trajectory.
The technology shift in the midstream can be rapid and dramatic, in particular when
it is linked to increasing urban demand, improving infrastructure, and a policy of
encouragement and support. The case of the rapid rise of potato cold storages in Bihar
(Minten et al., 2014) and Western Uttar Pradesh near Delhi (Das Gupta etal., 2010)
illustrate. We focus on the latter. Asurvey of cold storages in Agra found that the com-
bination of the rapid development of vegetable demand in Delhi, the improvement of
the road link from Agra to Delhi, the introduction of a disease-resistant and long-shelf-
life potato variety, the introduction of an electricity grid, the partial subsidizing of
irrigation pumps and cold storage equipment, and the economy’s generating investable
funds among the intermediate city business sector, led to very rapid and deep change
in the cold storage sector in Agra and, in turn, on the seasonality and cost of potatoes
in Delhi and intermediation patterns in the rural area. In the early 1990s relatively few
farmers grew potatoes in Agra and there were almost no modern cold storages. By the
late 1990s cold storages had risen to store 40 per cent of the vastly larger potato out-
put, and by 2009, 80 per cent. Traditional on-farm storage went from ubiquitous to 1
per cent of the potato harvest. Delhi went from sharply seasonal potato consumption
(from fresh harvest) to multi-season availability and 65 per cent of consumption from
cold-storage potatoes mainly from Agra. Rural brokers were sidelined by the cold stor-
ages themselves becoming the main locus of intermediation with urban wholesalers
coming to buy potatoes from farmers at the storages.
(v) Conduct change: changing financial relationships between
farmers and retailers
A key policy assumption about the midstream segment in Asia has been the traditional
view that ‘tied output-credit markets’ are the ubiquitous way in which traders deal with,

ThomasReardon
58
and in a sense entrap and exploit, farmers. This translates to the idea that traders practice
of value chain nance, where a trader advances funds to a farmer, extracting the prom-
ise from the farmer that he/she will sell the crop to that trader at harvest; it is usually
hypothesized that hidden in that advance is a high interest rate, and the entrapment of
the farmer so that he/she must accept a low crop price. The idea is that the credit market
is missing so that farmers are forced to look to the trader for credit to make ends meet
until the harvest, and is thus willing to enter a ‘tied output-credit market’ arrangement.
These tied arrangements were often the justication for grain parastatals to sideline
exploitative traders and for governments to develop agrarian banks to resolve miss-
ing credit markets. During the ‘structural adjustment decades’ of the 1980s–2000s, the
hypothesis of the existence and the dominance of traditional value chain nance was
not, or very rarely, questioned, and not explored empirically. In fact, one nds masses
of attestations to the persistence of the belief in the dominance of this traditional
nance in policy debates and academic papers in the past decade. This persistence is
important because it casts doubt on whether farmers can gain from the development of
the midstream segment of the value chain and the ‘market’ in general, and whether the
market is really ‘free’ even when it is de jure ‘liberalized’, as the assumption is that these
tied arrangements persist.
These ideas have recently been re-examined by Reardon etal. (2014c) using survey
data on potato and rice farming in China, India, Bangladesh, and Vietnam. They nd
there is little value chain nance ‘upstream’ in the rice or potato value chain in Asia, in
both more and less developed areas. But there is a lot of value chain nance midstream
and downstream; however, these intra-midstream (i.e. between two sub-segments of the
midstream, such as between processors and wholesalers) or between midstream and
downstream (with processors and/or wholesalers on the midstream side, and retailers
on the downstream (segment) side) advances and delayed payments are frequent but
minor, as they usually involve tight transaction times of a week or so, just the churning
of money in the regular relationships and networks of traders and clients and suppliers.
(vi) Conduct change: incipient emergence of market institutions in
the form of contracts and private standards
While structural and technological changes in the midstream are more widespread and
advanced, there have nonetheless been some important institutional and organizational
changes in agricultural value chains, although these have been concentrated in the ini-
tial, emergent stage. They tend to be correlated with the waves of transformation over
the countries and products, mainly linked to large and especially multinational compa-
nies, and related more to products that are highly perishable, that present potential food
safety and thus liability issues. Of course, given that the consumption of perishables,
and the role of larger companies is rising, these institutional mechanisms will gradually
take on an increasing importance and role. For now, we discuss them briey as follows.
First, there is some evidence that there has been an incipient shift from solely spot
markets to contracts. There is no systematic information on how advanced is the diffu-
sion of contracts among segments in value chains in developing Asia; contract farming
linked to processors seems to be emergent only in a few product categories and countries,
such as for some export vegetables and fruit, and some processed vegetables, such as

59
The midstream of agrifood value chains in developing countries
potatoes for French fry production by large companies, or feed grain operations, such as
maize in Indonesia (Simmons etal., 2005); in pork and chicken (for example in Thailand;
NaRanong, 2008) and milk in some countries and only so far as an emerging segment.
In staples, such as rice, there is some anecdotal evidence of contract farming but surveys
tend to nd little evidence of it. Reardon etal. (2014b) reported that in interviews large
mills claimed that they were using contracts but farm surveys in their catchment areas
showed very few to no contract arrangements with farmers. They nd no evidence of
contracts in domestic fruit and vegetable or sh value chains, apart from in the export of
shrimp or other export-oriented arrangements by multinational companies.
Second, there has been a shift from no public standards, to the emergence of public
health standards and the emergence of private standards. Again, there is no systematic
assessment of the diffusion of private standards. There appears to be a correlation
between how traded a product is into the international market, and/or its degree of pro-
cessing, and the degree of public standard emergence for the product. Private standards
tend to be for internal procurement across borders of multinational retailers, and for
some large processing companies such as Nestlé sourcing from an array of countries
and selling in the regional market.
Third and more common is the shift from no brands to the spread of branding for
packaged and processed foods, and especially in the more advanced parts of develop-
ing Asia. Reardon etal. (2014b) analyse the evidence from rice retail surveys in urban
areas of China, Bangladesh, India, and Vietnam to show that there has been a relatively
recent but rapid shift from sale of loose or poly-packed unbranded rice to mill-branded
rice in China, and to a lesser extent in India and Bangladesh.
Fourth, as noted above, especially for large retail chains and large processors, there
has been emerging organizational change (especially for dry and frozen processed
foods; less so for fresh fruits and vegetables, so far)—the emergence of centralized pro-
curement systems of large food industry companies with use of distribution centres,
regional and global networks, and specialized dedicated wholesalers.
IV. Conclusions
The paper shows that there has indeed been rapid growth and transformation of the
midstream segments, both in a ‘modern revolution’ with the ingress of large and often
foreign companies, but also a ‘quiet revolution’ with a proliferation of SMEs and sub-
stantial investment by them. These revolutions have been spurred at rst by direct
government action, but then after liberalization and privatization, and after a take-off
of urbanization, income growth, and diet change, and vast improvement in hard and
sometimes in soft infrastructure, by private-sector investment, overwhelmingly impor-
tant compared to direct government investments, and for the domestic market, over-
whelmingly important compared to the internationally traded sectors.
During these processes, agrifood policies have played important roles. Sometimes
the roles have been to slow transformation, such as in the market limiting and ‘reserv-
ing’ regulations of India, or in the lack of public investment in upgrading wholesale
markets, such as in Indonesia. But often policy and public investment have spurred
transformation of the midstream—by liberalizing FDI which had a much larger role

ThomasReardon
60
than product trade in changing the Asian food value chain midstream; by putting in
place ‘enabling conditions’, such as roads and electricity to make protable the private
investments; by sometimes improving upstream supply conditions to the midstream,
such as in the introduction of potato varieties that are more storable and shippable;
by putting in place commercial regulations that improved the domestic ‘food business
climate’, and occasionally by subsidizing equipment and plant investments to upgrade
processing and logistics.
References
Ablett, J., Baijal, A., Beinhocker, E., Bose, A., Farrell, D., Gersch, U., Greenberg, E., Gupta, S., Gupta,
S. (2007), The ‘Bird of Gold’: The Rise of India’s Consumer Market, San Francisco, McKinsey
Global Institute.
Abrami, R. M., Kirby, W. C., McFarlan, F. W., and Yuen Manty, T. (2008), ‘Inner Mongolia Yili
Group: China’s Pioneering Dairy Brand’, Harvard Business School Case 9-308-052.
Ahmadi-Esfahani F. Z., and Locke, C. G. (1998), ‘Wholesale Food Markets with ‘Chinese characteris-
tics’’, Food Policy, 23(1), 89–103.
Anderson, K., Dimaranan, B., Hertel, T., and Martin, W. (1997), ‘Asia-Pacic Food Markets and
Trade in 2005: AGlobal, Economy-wide Perspective’, The Australian Journal of Agricultural and
Resource Economics, 41(1), 19–44.
Bell, D. E., and Shelman, M. (2010), ‘Yum! China’, Harvard Business School Case 9-511-040,
December.
Bennett, M. K. (1954), The World’s Food, New York, Harper.
Bhavani, T. A., Gulati, A., and Roy, D. (2006), ‘Structure of the Indian Food Processing Industry: Have
Reforms Made a Difference?’, presented at the workshop, ‘From Plate to Plough: Agricultural
Diversication and its Implications for the Smallholders’, New Delhi.
Byerlee, D. (2014), ‘The Fall and Rise Again of Plantations in Tropical Asia: History Repeated?’, Land,
3, 574–97.
— Kyaw, D., San Thein, U., Seng Kham, L. (2014), ‘Agribusiness Models for Inclusive Growth in
Myanmar: Diagnosis and ways forward’, report to USAID by Michigan State University and the
Myanmar Development Research Institute.
Collinson, S., and Rugman, A. M. (2007), ‘The Regional Character of Asian Multinational Enterprises’,
Asia Pacic Journal of Management, 24, 429–46.
Das Gupta, S., Reardon, T., Minten, B., Singh, S. (2010), ‘The Transforming Potato Value Chain
in India: From a Commercialized-Agriculture Zone (Agra) to Delhi’, Report of Value Chains
Component of Asian Development Bank RETA (13th) IFPRI Project on Policies For Ensuring
Food Security In South And Southeast Asia, October.
Deichmann, U., Shilpi, F., and Vakis, R. (2009), ‘Urban Proximity, Agricultural Potential and Rural
Non-farm Employment: Evidence from Bangladesh’, World Development, 37, 645–60.
Deininger, K., and Feder, G. (2001), ‘Land Institutions and Land Markets’, Handbook of Agricultural
Economics, Amsterdam, Elsevier.
Dolan, C., and Humphrey, J. (2000), ‘Governance and Trade in Fresh Vegetables: The Impact of
UK Supermarkets on the African Horticulture Industry’, Journal of Development Studies, 37(2),
147–76.
Fafchamps, M., and Shilpi, F. (2003), ‘The Spatial Division of Labour in Nepal’, Journal of Development
Studies, 39(6), 23–66.
Fan, S., and Chan-Kang, C. (2005), ‘Road Development, Economic Growth, and Poverty Reduction
in China’, Research Report 138, Washington, IFPRI.

61
The midstream of agrifood value chains in developing countries
Feller, A., Shunk, D., and Callarman, T. (2006), ‘Value Chains versus Supply Chains’, BP Trends,
March, accessed at www.bptrends.com 23 April 2015.
Figuié, A., Pham, T., and Moustier, P. (2013), ‘Grippe aviaire dans la lière: La réorganisation du
secteur agro-industriel au Vietnam’, Revue d’Etudes en Agriculture et Environnement, 94(4),
397–419.
Gorton, M., Sauer, J., and Supatpongkul, P. (2011), ‘Wet Markets, Supermarkets and the ‘Big Middle’
for Food Retailing in Developing Countries: Evidence from Thailand’, World Development, 39(9),
1624–37.
Haggblade, S., Hazell, P., and Reardon, T. (eds) (2007), Transforming the Rural Nonfarm Economy:
Opportunities and Threats in the Developing World, Baltimore, MD, Johns Hopkins University
Press.
Hollinger, F., and Staatz, J. (2014), Agricultural Growth in West Africa: Market and Policy Drivers,
Rome, FAO.
Huang, J., Huang, Z., Zhi, H., Wu, Y., Niu, X., and Rozelle, S. (2007), Production, Marketing and
Impacts of Market Chain Changes on Farmers in China: Case Study of Cucumber and Tomato in
Shandong Province, Micro Study Report of Component 1 of the Regoverning Markets Program,
Beijing, Chinese Academy of Science. September.
Kumar, A., Parappurathu, S., and Joshi, P. K. (2013), ‘Structural Transformation in Dairy Sector of
India’, Agricultural Economics Research Review, 26(2), 209–19.
Lele, U. (1971), Grain Marketing in India: Private Performance and Public Policy, New York, Cornell
University Press.
Levenstein, H. (1988), Revolution at the Table, New York, Oxford University Press
McLeod, A., Thieme, O., and Mack, S. D. (2009), ‘Structural Changes in the Poultry Sector: Will
There be Smallholder Poultry Development in 2030?’, World’s Poultry Science Journal, 65(June),
191–200.
Mellor, J. W. (1976), The New Economics of Growth, Ithaca, NY, Cornell University Press.
Minten, B., Reardon, T., and Sutradhar, R. (2010), ‘Food Prices and Modern Retail: The Case of
Delhi’, World Development, 38(12), 1775–87.
— Singh, K. M., Sutradhar, R. (2014), ‘The New and Changing Roles of Cold Storages in the Potato
Supply Chain in Bihar’, Economic and Political Weekly, 49(52), 98–108.
Morisset, M., and Kumar, P. (2008), ‘Structure and Performance of the Food Processing Industry in
India’, Laval, mimeo.
Moustier, P. (2009), ‘Governance and Performance of Food Chains in Vietnam’, Economie et Sociétés—
Cahiers de L’ISMEA, Série ‘Systèmes Agroalimentaires’, no.11, November, 1835–57.
NaRanong, V. (2008), ‘Structural Changes in Thailand’s Poultry Sector: Avian Inuenza and its
Aftermath’, TDRI Quarterly Review, 23(3), 3–10.
Natawidjaja, R., Reardon, T., and Shetty, S., with T. I. Noor, T. Perdana, E. Rasmikayati, S. Bachri,
and R. Hernandez. (2007), Horticultural Producers and Supermarket Development in Indonesia,
UNPAD/MSU/World Bank, World Bank Report no. 38543, published by the World Bank/
Indonesia, July.
National Bureau of Statistics of China (2012), Annual Statistics, 2011, Beijing, Government of China.
Pingali, P. (2006), ‘Westernization of Asian Diets and the Transformation of Food Systems:
Implications for Research and Policy’, Food Policy, 32, 281–98.
— Rosegrant, M. W. (1995), ‘Agricultural Commercialization and Diversication: Processes and
Policies’, Food Policy, 20(3), 171–85
Popkin, B. M. (2014), ‘Nutrition, Agriculture and the Global Food System in Low and Middle Income
Countries’, Food Policy, 47, 91–6.
Pray, C., and Naseem, A. (2007), ‘Supplying Crop Biotechnology to the Poor: Opportunities and
Constraints’, Journal of Development Studies, 43(1), 192–217.
Rashid, S., Gulati, A., and Cummings, R. (2008), From Parastatals to Private Trade: Lessons from
Asian Agriculture, Baltimore, MD, Johns Hopkins University Press.
Reardon, T., and Berdegué, J. A. (2002), ‘The Rapid Rise of Supermarkets in Latin America: Challenges
and Opportunities for Development’, Development Policy Review, 20(4), 317–34.

ThomasReardon
62
Reardon, T., and Timmer, C. P. (2014), ‘Five Inter-linked Transformations in the Asian Agrifood
Economy: Food Security Implications’, Global Food Security, 3(2), 108–17.
— Henson, S., and Berdegué, J. (2007a), ‘‘Proactive Fast-Tracking’ Diffusion of Supermarkets in
Developing Countries: Implications for Market Institutions and Trade’, Journal of Economic
Geography, 7(4), 1–33.
— Stamoulis, K., and Pingali, P. (2007b), ‘Rural Nonfarm Employment in Developing Countries in
an Era of Globalization’, Agricultural Economics, 37(s1), 173–83.
— Chen, K. Z., Minten, B., and Adriano, L. (2012), The Quiet Revolution in Staple Food Value Chains
in Asia: Enter the Dragon, the Elephant, and the Tiger, Asian Development Bank and IFPRI,
December.
— Zhang, K., and Hu, D. (2010), The Transforming Rice Value Chain in China: The Rice Road
from Heilongjiang to Beijing, Report of Value Chains Component of Asian Development Bank
RETA (13th) IFPRI Project on Policies For Ensuring Food Security in South and Southeast Asia.
November.
— Timmer, C. P., Barrett, C. B., and Berdegue, J. (2003), ‘The Rise of Supermarkets in Africa, Asia,
and Latin America’, American Journal of Agricultural Economics, 85(5), 1140–6.
— Tschirley, D., Minten, B., Haggblade, S., Timmer, C. P., Liverpool-Tasie, S. (2013), ‘The Emerging
‘Quiet Revolution’ in African Agrifood Systems’, brief for ‘Harnessing Innovation for African
Agriculture and Food Systems: Meeting Challenges and Designing for the 21st Century’; 25–26
November, African Union Conference Center, Addis Ababa, Ethiopia (Ko Annan Foundation
and B&M Gates Foundation).
— Dolislager, M., Snyder, J., Hu, C., and White, S. (2014a), Urbanization, Diet Change, and
Transformation of Food Supply Chains in Asia, report for USAID Michigan State University
Project of the Global Center for Food System Innovation and the Food Security Policy Innovation
Lab, May.
— Chen, K. Z., Minten, B., Adriano, L., Dao, T. A., Wang, J., Das Gupta, S. (2014b), ‘The Quiet
Revolution in Asia’s Rice Value Chains’, Annals of the New York Academy of Sciences, 1331,
106–18.
— Minten, B., Chen, K., Das Gupta, S., Dao, T. A., Wang, J., and Murshid, K. A.S. (2014c), ‘Tied
Output-Credit Markets Have Come Untied: The Fall of Traditional Agrifood Value Chain Finance
in Asia’, Paper for the Asian Development Bank, MSU, IFPRI, Manila, August.
Roemer, M., and Jones, C. (1991), Markets in Developing Countries—Black, Fragmented and
Parallel, Cambridge, MA, International Center for Economic Growth and Harvard Institute for
International Development.
Rosegrant, M. W., Gazmuri Schleyer, R., and Yadav, S. N. (1995), ‘Water Policy for Efcient
Agricultural Diversication: Market-based Approaches’, Food Policy, 20(3), June, 203–23.
Rushton, J., Viscarra, R., Guerne Bleich, E., and McLeod, A. (2005), ‘Impact of Avian Inuenza
Outbreaks in the Poultry Sectors of Five South East Asian Countries (Cambodia, Indonesia,
Lao PDR, Thailand, Viet Nam): Outbreak Costs, Responses and Potential Long Term Control’,
World’s Poultry Science Journal, 61(September), 491–514.
Schertz, L. P., and Daft, L. M. (eds) (1994), Food and Agricultural Markets: The Quiet Revolution,
Washington, National Planning Association, Economic Research Service, USDA.
Schneider, M. (2011), Feeding China’s Pigs: Implications for the Environment, China’s Smallholder
Farmers and Food Security, Minneapolis, Institute for Agriculture and Trade Policy.
Simmons, P., Winters, P., and Patrick, I. (2005), ‘An Analysis of Contract Farming in East Java, Bali,
and Lombok, Indonesia’, Agricultural Economics, 33, supplement, 513–25.
Skinner, G. W. (1964), ‘Marketing and Social Structure in Rural China: Part 1’, Journal of Asian
Studies, 24(1), 3–43.
Times of India (2014), ‘FDI in Food Processing Touches $2.15 billion’, Times of India, 23 January,
accessed from www.timesondia.indiatimes.com/business/india-business/ on 24 January 2015.
Timmer, C. P. (1973), ‘Choice of Technique in Rice Milling on Java’, Bulletin of Indonesian Economic
Studies, 9(2), 57–76.

63
The midstream of agrifood value chains in developing countries
Timmer, C. P. (2015), ‘The Dynamics of Agricultural Development and Food Security in Southeast
Asia: Historical Continuity and Rapid Change’, in I. Coxhead (ed.), Routledge Handbook of
Southeast Asian Economics, London and New York, Routledge, 89–113.
— Dawe, D. (2010), ‘Food Crises Past, Present (and Future?): Will We Ever Learn?’, in D. Dawe (ed.),
The Rice Crisis: Markets, Policies and Food Security, London, FAO and Earthscan.
Traill, W. B. (2006), ‘The Rapid Rise of Supermarkets?’, Development Policy Review, 24(2), 163–74.
Tschirley, D., Reardon, T., Dolislager, M., and Snyder, J. (2015), ‘The Rise of a Middle Classin Urban
and Rural East and Southern Africa: Implications for Food System Transformation’, Journal of
International Development, forthcoming, May.
Von Thünen, J. H. (1826), Der isolierte staat in Beziehung auf Landwirtschaft und Nationalökonomie,
Jena, Gustav Fischer.
Wang, J., Reardon, T., Chen, K., and Huang, Z. (2013), The Rice VC from Jiangxi to Zhejiang in China,
Report for Asian Development Bank TA-7648 REG; Regional—Research and Development
Technical Assistance (R-RDTA).
Yi, D., Reardon, T., and Stringer, R. (2015), ‘Modern Variety Adoption and Intensication in
Indonesian Shrimp Aquaculture: Are Small Farmers Included?’, Working Paper, University of
Adelaide and Michigan State University.
Xia, T. (2015), ‘US Implications of the Smitheld Acquisition by Shuanghui’, Choices, 30(1), 1–5.