Int. J.

Production Economics 168 (2015) 267278

Contents lists available at ScienceDirect

Int. J. Production Economics

journal homepage: www.elsevier.com/locate/ijpe

Changing a multidomestic production network to a global function

network: North America Heinz ketchup from 1960 to 2015
John Miltenburg
School of Business, DSB-411, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4M4

art ic l e i nf o a b s t r a c t

Article history: Two frameworks, the global production network and manufacturing strategy, are used to analyze the
Received 22 February 2015 North America Heinz ketchup production network during the time period from 1960 to 2015. The
Accepted 1 July 2015 frameworks reveal how and why Heinz changed its traditional multidomestic network to a modern
Available online 10 July 2015
global function network. This network enabled Heinz to become the industry's lowest-cost producer
Keywords: while producing the industry's leading product. It is also the network-type Heinz is establishing in its
Global production networks Europe, Asia/Pacic, and Rest of World business units. The analysis gives insight into the usefulness of
Operations strategy the two frameworks for analyzing complex production networks.
Case study: Heinz & 2015 Elsevier B.V. All rights reserved.

1. Introduction The rest of the paper is organized as follows. Section 2 reviews

This paper uses two frameworks to describe and analyze the
North America production network for Heinz ketchup from 1960
to 2015. From its beginning Heinz operated a multidomestic
production network. Then in the mid-1990s there was a ten-year
period of enormous, sustained pressure to reduce cost. Heinz
responded by changing its strategy and transforming its network
to a global function production network. Today Heinz produces the
industry's leading product and is the industry's lowest-cost
producer. These accomplishments are usually conicting; it is
usually not possible to be the lowest-cost producer while produ-
cing the best product. Accomplishing both gave Heinz an enor-
mous competitive advantage. It also made Heinz an attractive
take-over target. In February 2013 Berkshire Hathaway Inc. and 3G
Capital acquired Heinz in a friendly take-over.
Since the ownership change, Heinz has reduced the informa-
tion it makes public. It is now more difcult than ever to learn the
details of Heinz's production activities. It is timely, therefore, to
give a careful description and analysis of the production network
for Heinz's most important product family, Heinz ketchup. Two
frameworks are used: the global production network framework,
developed by economic geographers, and the manufacturing
strategy framework. The frameworks reveal the changes that were
made and the reasons they were made. The analysis gives insight
into (i) how a company becomes the lowest-cost producer while
producing the industry's leading product, and (ii) the usefulness of
the two frameworks for studying complex production networks.
E-mail address: miltenb@mcmaster.ca
the two frameworks, links them together, summarizes their
application to the Heinz production network, describes the meth-
odology used, and positions the analysis in the literature. Section 3
describes the H.J. Heinz Company and its North American products
and production activities. Section 4 examines the ketchup produc-
tion network from the company's early years to about 1993, from
1993 to 2005, and from 2005 to the present time. Section 5
discusses strategic implications. Section 6 gives conclusions and
suggestions for future research.
2. Two frameworks for analyzing production networks
Various frameworks are used to analyze production networks.
Friedli et al. (2014) give an over-review; see also Weir et al. (2000),
Miltenburg (2005), Rudberg and West (2008), and Cheng et al.
(2011). Two frameworks are used here: the global production
network framework, and the manufacturing strategy framework.
The global production network was developed by economic
geographers (Henderson et al., 2002; Dicken, 2003) and is a
familiar framework for social scientists (Reimer, 2007; Coe et al.,
2008). It is used in this paper to analyze the social forces shaping
the Heinz production network. The manufacturing strategy frame-
work is then used to analyze the detailed design of the network.
2.1. Global production network (GPN)
Fig. 1 displays the GPN framework. The bottom of the gure
shows three foundations: value, power, and embeddedness, on
which the production network rests. The network is further

http://dx.doi.org/10.1016/j.ijpe.2015.07.003
0925-5273/& 2015 Elsevier B.V. All rights reserved.
268 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278
Fig. 1. Global Production Network (GPN) framework.
shaped by three forces: the (industrial) sector in which a company
operates, external institutions impacting the network, and how
power is exercised through governance and ownership structures.
2.1.1. Network foundations
A facility is appropriate for a production network when the
conditions for creating, enhancing, and capturing value exist. For
ketchup the conditions are: tomato growers, low cost labor,
process technology, and transportation. Power can be corporative,
collective or institutional. Heinz, like many other companies, is
very centralized; all corporate power resides at head-ofce. Head-
ofce assigns production tasks and makes resource decisions.
Individual facilities have no corporate power. Grower associations,
labor unions, and co-processors have almost no collective power.
Large retail store chains, which emerged in 2000, have collective
power. Governments have a small amount of institutional power. A
facility is embedded in a geographic location where it absorbs and
becomes constrained by the social and economic activities that
exist there. From the early years of the company, Heinz became
more and more territorially embedded in three geographic loca-
tions. With great difculty it started to break out of this embedd-
edness in the late 1990s.
2.1.2. Network forces
Three forces shape the production network: sector, institutions,
and governance and ownership. Companies in the same industrial
sector use similar technologies and suppliers, produce similar
products for similar markets, and are affected by similar events.
Institutions such as unions, trade associations (e.g. the California
Tomato Growers Association, the Ohio Produce Growers and
Marketers Association), and governments all inuence the design
and operation of the production network. Governance and owner-
ship structures also shape the network.
2.2. Manufacturing strategy framework
The GPN framework ensures that production network design
takes account of the three foundations and three forces. The
detailed network design is then done in the manufacturing
strategy framework (Fig. 2). Two inter-dependent parts comprise
the framework: facility-type, and network-type.
2.2.1. Facility-type
Ferdows (1997) described the six facility-types: server, outpost,
offshore, contributor, lead, and source. While other categorizations
are possible (Vokurka and Davis, 2004) this categorization is well-
known (Vereecke and vanDierdonck, 2002; Maritan et al., 2004;
Feldmann and Olhager, 2013). Three characteristics help distin-
guish the six types: strategic purpose, scope of activities, and level
of capability (Fig. 2). The strategic purpose for a facility can be
(i) proximity to markets, (ii) access to process and product
technology, and (iii) access to factors necessary for low-cost
production. The scope of activities in a facility ranges from narrow
to broad. A facility with a narrow scope of activities only engages
in production activities. A facility with a broad scope of activities
also engages in some of: sourcing, distribution, design and
improvement, and research and development. The level of cap-
ability of a facility ranges from low to high. Many years of
production, improvement, and investment raise the level of cap-
ability from low to high. Level of capability and scope of activities
operate in parallel: a facility-type with a broad scope of activities
requires a high level of capability. Three facility-types (server,
outpost, and offshore) have a narrow scope of activities and a low
level of capability. They are predominant in simple production
networks. The other three facility-types (contributor, lead, and
source) have a broad scope of activities and a high level of
capability. They are needed in complex production networks.
2.2.2. Network-type
The network-types in Fig. 2 follow Shi and Gregorys (1998)
categorization. Other categorizations are possible (Afonso and
Fleury, 2009; Jaehne et al., 2009). There are nine network-types:
domestic, domestic export, multidomestic, international, multi-
national, global (function, product, mixed), and transnational. Two
characteristics help distinguish the types: pressure for local
responsiveness and pressure for globalization. Each pressure
ranges from low to high. Pressure for local responsiveness is low
when conditions are similar in all regions where a company
operates. When conditions vary from region to region, pressure
for local responsiveness is higher and the company must operate
differently in different regions. Pressure for globalization is high
when there are powerful customers, strong competitors and
suppliers, low tariffs, free trade regions, and so on (Porter, 1985).
When these are absent pressure for globalization is low.
The arrows in the gure linking facility-type and network-type
show inter-dependency (Colotla et al., 2003) between facility- and
network-types, and that facility location depends on network-type
(Hammamia et al., 2008). The arrows indicate that when pressure
for globalization is not high it is sufcient to use low capability,
local or regional facility-types. When pressure for globalization is
high, companies need high capability, dispersed facility-types.
2.3. Production network for Heinz ketchup in North America from
1960 to 2015
The next sections examine the North America production
network for Heinz ketchup during three time periods: the early
years to about 1993, 1993 to 2005, and 2005 to the present time.
As depicted in Fig. 3a, we will see that value creation (i.e. growing
tomatoes, producing tomato paste, and producing ketchup), power
(corporate head-ofce and institutional large retail-store chains),
embeddedness (in the primary North America tomato growing
areas), sector technology (large-scale tomato paste production and
storage), and ownership (acquisition by Berkshire and 3G) are the
important foundations and forces that shaped the production
network. As depicted in Fig. 3b, we will see that in the early years
to about 1993 Heinz used a multidomestic production network of
low capability facilities, from 1993 to 2005 Heinz made major
changes to its production network, from 2005 to the present time
Heinz used a global function production network of high capability
facilities, and indications today are that the network is evolving
into a global mixed network.
 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278 269

Fig. 2. Manufacturing strategy framework.

Future network:
beyond 2015

Current network:
2005 to present time

Change:
1993 to 2005

Early network:
early years to 1993

Fig. 3. (a). Foundations and forces shaping the Heinz production network. (b). Detailed design of the Heinz production network.

How companies change production networks to adapt to 2.4. Methodology

changing conditions is an important subject in the literature;
see, for example, Avella et al. (2001), Demeter (2003), and
Cagliano et al. (2005). Muehlfeld et al. (2011) analyze how
companies in the food industry, like Heinz, do this through
mergers and acquisitions. Garrido-Vega et al. (2015) examine the
dependency between production strategy and technology in the
achievement of competitive advantage in several industries.
Sydow and Frenkel (2013) examine how companies use a global
supply chain and exible labor to respond to risk and uncertainty.
The primary source data used in this paper comes from nancial
documents, news reports, and interviews from Heinz, other compa-
nies, and government and non-government agencies. The reports
and interviews used were published in print, on the internet, and on
television and radio by news agencies such as newspapers and
television news organizations. All data was triangulated from two
or more sources to verify its accuracy. Secondary source data was
taken from scholarly articles in the literature.
270 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278

3. H.J. Heinz Company
Heinz with sales of $11.7 billion in 2012 was the 11th largest
food company in the world (Heinz Annual Report, 2012). Heinz
organizes its products into four categories: ketchup and sauces
(sales of $5.2 billion in 2012), meals and snacks ($4.5 billion),
infant and nutrition ($1.2 billion), and other ($0.7 billion). Heinz
has ve business units (Table 1): North America Consumer
Products (production and sales of products in all categories to
grocery channels in the United States, Canada and Mexico), North
America Foodservice (production and sales of products in all
categories plus customized products to commercial food chan-
nels), Europe (production and sales of products in all categories in
Europe), Asia/Pacic (production and sales of products in all
categories in Australia, New Zealand, India, Japan, China, Papua
New Guinea, South Korea, Indonesia, Vietnam and Singapore), and
Rest of World (production and sales of products in all categories in
selected countries in Africa, Latin America, and the Middle East).
Table 2 shows the annual return on Heinz stock. A particularly
low return in 2005 provoked a confrontation between manage-
ment and several important shareholders. This led to a change in
the Board of Directors, the launch of productivity initiatives, and
an increase in foreign acquisitions. The table also contrasts the
return on Heinz stock with returns on: a sector indexthe
Standard and Poor Packaged Foods and Meats Index (S&P PFaM
Index), a general indexthe S&P 500 Index, and the change in
book value per-share of Berkshire Hathaway Inc. (the company
that would acquire Heinz in 2013). The average returns at the
bottom of the table show that Heinz slightly outperformed the
average company in its industry (the PFaM Index) and signicantly
outperformed the average company in the S&P 500. The rst result
suggests that Heinz protability could be improved (e.g. by a
change in strategy or management). The second result suggests
that Heinz is a protable company and, therefore, worth
Table 1
2012 sales and protability (Heinz Annual Report, 2012).
Business unit Sales ($ Operating income ($
billion) billion)
North America Consumer $3.242 $0.812
Products
North America $1.419 $0.166
Foodservice
Europe $3.441 $0.609
Asia/Pacic $2.569 $0.206
Rest of World $0.978 $0.105
Table 2
Annual returns from 2004 to 2012 (Heinz Annual Report, 2009, 2012; Berkshire
Hathaway Annual Report, 2014).
Year Return
considering for investment. The table also shows that the average
return at Berkshire is signicantly higher. For example, the 11.8%
average return on Heinz stock taxed at 40% gives an after-tax
return of 0.118
(1  0.4) 7.1%. This is signicantly less than the
9.8% average return at Berkshire. This suggests that Berkshire
could acquire Heinz if it expected that Heinz could earn a higher
return.
Heinz's 7.6% return in 2012 was good compared to large
companies in the S&P 500 index (5.8%), but poor compared to
companies in its sector (13.6%), and very poor compared to
Berkshire (14.4% after-tax). This made 2012 an opportune time
for Berkshire to try to acquire Heinz. In February 2013 Berkshire
Hathaway Inc. (a public holding company led by Warren Buffet)
and 3G Capital (a private Brazilian investment company) executed
a friendly take-over of Heinz for $72.50 per share or $23 billion (or
$28 billion if debt assumption is included). Berkshire and 3G each
owned one-half of Heinz, and 3G took over management of Heinz.
3G had the expertise to change Heinz strategy, drive improve-
ments, and, therefore, generate higher returns.
Table 3 lists 31 factories in the United States, Canada and
Mexico that comprised the Heinz North America production
facilities for all products in all categories. Seven factories were
closed before the 2013 ownership change. Three more were closed
after the change. In addition to the factories, Heinz headquarters
were located at One PPG Place (225 employees) and at Heinz 57
Center on Sixth Street (700 employees) in Pittsburgh, Pennsylva-
nia. The details in Table 3 were compiled from nancial documents
and news reports from Heinz, other companies, and news agen-
cies. Two or more sources were used for each detail in the table in
order to verify its accuracy. The details were difcult to compile
because after the ownership change, Heinz greatly reduced the
information it made public.
Fig. 4 locates the 31 factories on a map. Notice the large number
of factories in California (seven) and the Ohio area (six). Heinz is
embedded in these areas. Three of the seven factories that were
closed before 2013 are located in the south-eastern United States.
This indicates a strategic decision to leave this geographic area.
Production activities also ceased at two factories in the north-east
Gross and two factories in California. The closures indicate that even
margin (%) before the ownership change Heinz was making its network more
focused and less dispersed. The three factories closed after 2013
0.812/
3.242 25 continued this pattern.
12 North America Consumer Products is Heinz's largest (28% of
sales) and most protable (25% margin) business unit. And
18 ketchup is Heinz's most important product family: Heinz is
8
11
number-one in ketchup globally (Heinz Annual Report, 2012, p.
3). Therefore North America production and sales of Heinz
ketchup is a critical element in the company's strategy. Ever since
the late 1990s Heinz's strategic objectives for North America
ketchup were to produce the industry's best product and to be
the industry's lowest-cost producer. Usually these objectives are
incompatible. In what follows we examine how Heinz changed its
production network so that both objectives could be achieved.

Heinz stock S&P PFaM Index S&P 500 Index Berkshire Hathaway
per-share book value
4. Heinz ketchup
2004 0.314 0.286 0.246 0.105
2005  0.002 0.064 0.049 0.064 Tomatoes are the main ingredient in ketchup. The main
2006 0.175  0.024 0.153 0.184 growing areas for tomatoes are shown in dark red in Fig. 5. They
2007 0.149 0.205 0.165 0.110
are California and the Ohio area in the United States, adjacent
2008 0.043  0.018  0.055  0.096
2009  0.248  0.217  0.353 0.198 south-west Ontario in Canada, and Italy, Spain and Portugal, Iran,
2010 0.396 0.402 0.392 0.130 and Turkey in Europe. Global production of tomatoes was 160
2011 0.163 0.164 0.161 0.046 million tons in 2014. There are two types of tomatoes: fresh-
2012 0.076 0.136 0.058 0.144 market and processing. Fresh-market tomatoes are large (1016
Average 0.118 0.111 0.091 0.098
Before-tax After-tax
ounces), soft (they break and splash when dropped), and avorful.
They are grown outside, or inside in green houses. They are
 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278 271

Table 3
Heinz North America factories for all products in all categories.

Location and name Zip code Year Employment full- Primary products Other information
time
open close

1. Ohio and surrounding area

1a.Fremont 43420- 1937 open 525 Ketchup World's largest ketchup plant
8584
1b.Massillon 44647- ? open 700 Frozen, nutrition Frozen food center of excellence
7669
1c.Mason 45040- 1989 open 450 Condiment portions, sauces
9601
1d.Holland, Michigan 49423- 1897 open 200 Pickles, vinegar, soup
3445
1e.St. Marys, Ontario, Canada N4X 1B7 2004 open 200 Salad dressing, sauces 3 plants in Canada
1f.Etobicoke, Ontario, Canada M8Z 2L8 ? open ? Juice, foodservice products Leamington was Heinz's largest plant
a
1g.Leamington, Ontario, N8H 3W8 1909 2014 800 Ketchup, sauces, beans, etc.
Canada
2. California and surrounding area
2a.Chatsworth 91311- 1994 open 249 Sauces, condiments
5923
2b.Escalon 95320-  1950 open 150 Tomatoes: canned, sauces 400,000 t/yr of processed tomatoes for
9601 restaurants
2c.Irvine 92614- 2001 open ? Frozen soups, sauces, LP Todds Division
6250 condiments
2d.Phoenix, Arizona 85043- 2001 open ?
4622
2e.San Diego 92154- 2001 open 440 Frozen foods DeliMex (Mexican) foods
8305 1 plant in US
2f.Monterry, Nuevo Leon, CP 66493 2001 open 400 1 plant in Mexico
Mexico
b
2g.Tracy 95304-  1940 1997 ? Ketchup Production moved to 1a
9301
b
2h.Stockton 95206 ? 2011 140 Ketchup, canned tomatoes Production moved to 1a, 2b, 4b
2i.Guadalajara, Jalisco, Mexico CP 45580 2007 open ? Ketchup
3. Florida and surrounding area
3a.Jacksonville, Florida 32219-  1995 open 250 Condiment portions
5313
3b.Fort Meyers, Florida 33905- 1991 open 230 Frozen appetizers
5013
a
3c.Florence, South Carolina 29506-  2005 2014 200 Frozen Production moved to 1b
8233
b
3d.Stone Mountain, Georgia 30083- 1999 2008 ? Sauces, condiments
1109
4. Other areas
4a.Cedar Rapids, Iowa 52404- ? open 260 Frozen soup, stew, sauces Heinz Quality Chef Foods, Inc.
7601
4b.Muscatine, Iowa 52761- 1893 open 320 Soup, sauces Heinz's oldest plant; bulk foodservice products
4521
4c.Ontario, Oregon 97914- 1965 open 750 Frozen Ore-Ida potato-based frozen foods
2535
4d.Pennsauken, N. Jersey 08110- ? open 100 Sauces
2304
4e.West Chester, Penn. 19380- ? open ? Frozen lled pasta
4274
4f.Le Center, Minnesota 56057- 2002 open 200 Frozen deserts Alden Merrell & Dianne's Fine Deserts
1200
b
4g.Newburyport, Mass. 01950- 2000 2012 175
4015
a
4h.Pocatello, Idaho 83202- ? 2014 410 Frozen Production moved to 1b
1963
b
4i.King of Prussia, Penn. 19406- 1972 2011 ? Frozen soup
2424
b
4j.Nashville, Tennessee 37210- ? 2006 ? Sauces
2506
b
4k.Dallas, Texas 75243- 1989 2008 200 Sauces, condiments Production moved to 1c, 2a, 3a
5501

a
Three factories closed after 2013.
b
Seven factories closed before 2013.

handpicked often when green (and then are gassed to promote taste (avoring spices are added later). They are grown outside
ripening), and are sold on the open market. and are mechanically harvested when red and fully ripe. They are
Processing tomatoes are very different. They are small (about grown under contract to food processors (hence the name,
3 ounces), hard (they bounce when dropped), and have a bland processing tomatoes) who produce four main products: tomato
272 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278

Canada

United States

Mexico

seven factories closed before the 2013 ownership change

Fig. 4. Heinz North America factory locations for all products in all categories.

Fig. 5. Areas of the world where tomatoes are grown shown in dark-red (Tomato News Online, 2014). (For interpretation of the references to color in this gure legend,the
reader is referred to the web version of this article.)

paste, ketchup and sauce, canned tomatoes, and juice. One quarter Table 4
of global production (
160 40 million tons per year) are Quantities of processing tomato grown (2014) (World Processing Tomato
processing tomatoes. Heinz is the world's largest user of proces- Council, 2015).

sing tomatoes (2.5 million tons in 2010). Table 4 gives global North America California 12.6 million tons
statistics for processing tomatoes. Other US states 0.6
Processing tomato varieties are developed to maximize the Canada 0.5
conversion of sunlight into sugar and solids, to be resistant to Europe Italy 4.9
Spain 2.7
insects and disease so that fewer applications of pesticides are
Portugal 1.2
required, and to be eshy and have a thick skin so they can Iran 2.2
withstand a harvesting machine's rough handling and can be Turkey 1.8
transported without damage in large trailers. (One trailer holds Asia-Pacic China 6.3
50,000 tomatoes or 50,000 tomatoesCapproximately 6 tomatoes Thailand 0.3
Australia 0.2
per pound 30,000 pounds or 15 t of tomatoes.) They are also South America Brazil 1.4
developed so that all tomatoes on a plant and all plants in a eld Chile 0.8
ripen simultaneously so that an entire eld can be mechanically Argentina 0.4
harvested at one time. 40 t of tomatoes are grown on one acre and
a mechanical harvester picks one ton of tomatoes every 20 s. At
the processing factory, tomatoes are usually processed into tomato
paste and packaged in 3000-pound, sterile boxes. Harvesting, 4.1. Early years to mid-1990s: multidomestic production network
transportation, and processing must be completed within six
hours; so the processing factory must be located close to Table 5 describes the primary North America growing areas for
growers' elds processing tomatoes. Heinz is embedded in each of these areas.
 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278 273

California is the largest area. The greatest number of factories (six) In terms of the GPN framework: Heinz created value in these
in any state are located there (recall Table 3). The second largest areas and was embedded there. Corporate power resided in
growing area is the Ohio, Indiana and Michigan area (four another area, head ofce in Pittsburgh. Growers associations and
factories) followed closely by south-western Ontario (two fac- farmworker unions had very little collective power. Heinz con-
tories). All of Heinz's 31 factories use tomato paste as a raw tracted with growers as it saw t and paid what it wanted. In
material but only six produce ketchup, and all six are located in terms of the manufacturing strategy framework: the Heinz fac-
the three areas. Table 6 lists the six ketchup factories and Fig. 6 tories had low capability, had a narrow scope of activities, and had
locates them on a map. Three are in California, two are in the Ohio two strategic purposesto be within a short distance of growers
area, and one is in Ontario. and to provide low-cost processing. Pressure for local responsive-
Fig. 6 is illustrative of the Heinz strategy during this time ness and pressure for globalization were low. The numerous, small
period. Heinz contracted local farmers to grow specic varieties of factories comprised a geographically dispersed, loosely coordi-
tomatoes, transport them to a local factory where the tomatoes nated multidomestic production network.
were processed into tomato paste, and later into ketchup, sauces,
and other products. Tomatoes grown in California were processed
in the local factories: 2b.Escalon, 2g.Tracy, and 2h.Stockton. Toma- 4.2. Pressure to change the production network
toes grown in Ohio were processed in the local factory, 1a.Fremont.
Tomatoes grown in Indiana were processed in the local factory, 4b. In 1984 Heinz's market share was 45%, more than double the
Muscatine. Tomatoes grown in Ontario were processed in the local share of its nearest competitor, Hunts ketchup. Heinz maintained
factory, 1g.Leamington. So, for example, in 1983 when Heinz its market share and protability by raising prices whenever costs
needed more tomatoes for its factories in Fremont and Muscatine, increased. This changed in the 1990s. The market was mature, the
it contracted tomato growers in Kouts, Indiana (Fig. 6), exactly economy was emerging from recession, and there was a new
mid-way between the two factories to grow tomatoes and then competitive product, no-name ketchup. While brand name
ship them to whichever factory had need. ketchup (e.g. Heinz and Hunts) sold at higher prices, no-name
ketchup sold at a low price. In 1994 Heinz sales dropped and prot
fell 15%. Heinz responded by launching productivity initiatives to
reduce costs. The initiatives were successful, and by the late 1990s
Table 5
North America areas growing processing tomatoes (2012) (United States market share began to rise, reaching 60% in 2002.
Department of Agriculture, 2015; Statistics Canada, 2015). But pressure to reduce cost did not diminish. Two new sources
emerged between 1993 and 2005: private-label ketchup and large
Growing area Acres Amount grown (tons of Yield (tons/ Value ($ store chains. Beginning in about 1995 and continuing every year
planted tomatoes) acre) per ton)
thereafter, the Heinz annual report (see Heinz Security Exchange
California 258,000 11,640,000 45 75 Commission form 10-K, 2015) identied private-label products as
Ohio/Indiana/ 18,000 540,000 30 113 principal competition: The Company regards its principal com-
Michigan petition to be other manufacturers of processed foods, including
Ontario, 14,000 560,000 40 95
branded, retail products, foodservice products and private label
Canada
products (italics added). Private-label ketchup was popular with

Table 6
North America factories producing ketchup.

Factory Year Full-time employees Other information

Open Close

1a.Fremont, Ohio 1937 open 525 World's largest ketchup plant

4b.Muscatine, Iowa 1893 open 320 Heinz's oldest plant
2b.Escalon, California  1950 open 150 400,000 t/year of specialty processed tomatoes
2g.Tracy, California  1940 1997 ? Production moved to 1a
2h.Stockton, California ? 2011 140 Production moved to 1a, 2b, 4b
1g.Leamington, Ontario 1909 2014 800 Heinz's largest factory

Growing area: Ontario, Ohio,

Indiana, Michigan
Factories: 1a, 1g, 4b

Kouts, Indiana

Growing area: California

Factories: 2b, 2g, 2h

Fig. 6. North America ketchup factory locations.

274 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278
customers because the quality was high (no-name ketchup quality
was low) and the price was up to 30% lower than Heinz. For
example, when the retail price was $0.07 per ounce for Heinz
ketchup and $0.06 for Hunts then the price was $0.05 for private-
label brands. Private-label ketchup was also popular with stores
because the prot margin was high. The wholesale price of Heinz
ketchup was about 20% higher than private-label ketchup but the
retail price was not 20% higher; so the prot margin on Heinz
ketchup was low. In 2009 the Heinz annual report acknowledged
this for the rst time: Private label brands sold by retail
customers, which are typically sold at lower prices, are a source
of competition for certain of our product lines.
The second source of continuing pressure to reduce cost was
the emergence of large retail store chains beginning in 2001. From
1995 to 2000 the Heinz annual reports stated: The Company is
not dependent on any single customer or a few customers for a
material part of its sales. This changed in 2001 when Heinz stated
for the rst time its concern over consolidation of retail stores into
large chains: ability to maintain its prot margin in the face of
a consolidating retail environment. The next year Heinz had a
single customer that accounted for more than 10% of its sales: (f)
or Fiscal 2002, Wal-Mart Stores, Inc. represented more than 10% of
the Company's sales. Two years later in 2004 Heinz described this
pressure as broad and global: ability to maintain its prot
margin in the face of a consolidating retail environment and large
global customers. Finally in 2006 the annual report was very
specic: As the retail grocery trade continues to consolidate, the
larger retail customers of the Company could seek to use their
positions to improve their protability through lower pricing.
The Heinz annual reports also stated that Heinz would respond
to this pressure by using its brand recognition (i.e. Heinz ketchup
is the industry's best product), product innovation (i.e. Heinz
produces a very wide variety of bottle shapes and sizes), and scale
(i.e. Heinz is the industry's low-cost producer). But Heinz could
not achieve scale with its multidomestic network of numerous,
small, low capability factories. To achieve scale, Heinz needed a
global function network comprised of a small number of very
large, high capability factories. So from 1993 to 2005 Heinz started
the very difcult work of changing its production network.
4.3. 1993 to 2005: starting the upgrade to a global function
production network
In the early 1990s the multidomestic production network
ceased to be effective. Heinz determined that it needed to
(1) reduce the cost of tomatoes, (2) reduce the factory cost of
processing the tomatoes, and (3) adopt newly developed technol-
ogy that made possible large scale production, storage, and
distribution of tomato paste. Heinz determined that the best way
to accomplish these three tasks was to close many of its
low capability factories, upgrade the others to high capability
factories, and organize the factories in a global function production
network.
Table 7
Price paid to growers and grower productivity (United States Department of Agriculture, 2015; Statistics Canada, 2015).
Growing area Price
$ per ton of tomatoes grown
1993 2006 2009
California $50 $58 $80
Ohio, Indiana $63 $82 $98
Ontario (Canada) $75 $75 $94
4.3.1. Task 1: reduce the cost of tomatoes
When Heinz pays growers $80 per ton of tomatoes, which is
$0.04 per pound, the cost of tomatoes is $0.04C3 $0.013 per
ounce of ketchup (because one pound of tomatoes produces
approximately three ounces of ketchup). If retail customers pay
$0.07 per ounce of Heinz ketchup then $0.013C$0.07 19% of the
price of ketchup is the cost of growing tomatoes. This high percent
means Heinz has no choice but to pay tomato growers the lowest
possible price.
Heinz contracts growers to grow specic quantities of specic
tomato varieties with specic planting and harvest dates. Growers
hire workers (usually temporary, foreign workers) to prepare and
harvest the crop. In this three-agent structureHeinz, growers,
workerseach agent is an independent contractor. This means
there are no employeremployee relationships and, therefore,
government regulations on working conditions, minimum wages,
and health and safety, which pertain only to employers and
employees, do not apply. Each worker participates in preparing
the tomato crop for several weeks and then participates in
harvesting the crop for several more weeks. At the end of the
harvest, the worker is paid a share of the crop. Most of the time,
this share is less than the government minimum hourly wage. To
ensure a successful harvest workers work long hours often
assisted by spouses and children. Workers often live in poor
housing and have little protection from pesticide exposure or
other unsafe working conditions.
In 1987 after a long and difcult process Heinz signed a collective
agreement in Ohio with a union called the Farm Labor Organizing
Committee (New York Times, 1987; Nichols, 1987). The agreement
established (i) an employeremployee relationship between growers
and about 2000 workers, and (ii) a relationship between growers and
Heinz. Growers were responsible for working and living conditions on
their farms, and Heinz was responsible for providing growers with
resources to pay the negotiated wages. Farms in California and Ontario
were not covered by the agreement. The agreement increased the cost
of tomatoes from the Ohio farms; so six years later in 1993 Heinz
stopped sourcing tomatoes from Ohio, Michigan, and Indiana in favor
of sourcing tomatoes from California. Tomatoes for Canadian produc-
tion continued to be sourced from Ontario growers until 2009 when a
dispute arose there over pricing. An arbitrator settled the dispute in
favor of the growers. This increased the cost of tomatoes so Heinz
immediately reduced its sourcing (and production) by 30% and moved
that work to California. A few years later in 2014, it stopped all sourcing
of tomatoes from Ontario and closed its 1g.Leamington, Ontario factory.
In California workers were employed by farm labor contractors (FLCs)
who contract with growers to provide low-cost workers. FLCs look
after immigration, housing, transportation, etc. and are not affected by
union agreements. This different three-agent structureHeinz,
growers, and FLCshas kept the cost of California tomatoes very low.
From 2001 to 2005, California growers were paid about $50 per ton
of tomatoes. In 2006, the price increased to $58 per ton (Table 7)
because of increases in the cost of energy, fertilizer, pesticide, water, and
seeds (Raine, 2007). Cost increases were partially offset by productivity
increases. In 1983 California growers harvested 26 t of tomatoes per
Productivity
Tons of tomatoes harvested per acre
2012 1983 1992 2005 2009
$75 26 33 40 43
$103 ? 25 26 31
$95 18 24 39.6 40.8
 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278 275

Table 8 planting in mid-March and the regular planting in mid-April.

California tomato paste producers2014 factories and capacities (Morning Star Harvest begins in mid-July and ends in late October giving a 3
Packing, 2014).
month harvest. In Ohio, Michigan, Indiana and Ontario there is no
Company Factory Year Capacity (tons per early planting and harvest begins in mid-August and nishes in
location built hour) mid-September, which is a one month harvest. This means the
tomato processing operation (which is separate from the ketchup
Type Name Tomatoes Tomato production operation) works for 3 months in California com-
paste
pared to one month in Ohio, Michigan, Indiana and Ontario. The
Private Morning Star Los Banos 1990 661 108 tomato processing operation is capital intensive; so utilizing the
Packing operation for 3 months is much more cost effective than utilizing
Williams 1995 749 123 it for only one month.
Liberty Packing Santa 1975 880 144
In 1997 the efciency of the tomato processing operation at the
Nella
Olam Tomato Lemorre 1990 448 73 1a.Fremont, Ohio factory was low. So Heinz closed the operation
Processors and moved that work to California. Because there was no tomato
Williams 1982 249 41 processing operation, Heinz also stopped sourcing tomatoes from
Average 587 98 Ohio, Michigan and Indiana. The efciency of the ketchup produc-
Grower-owned Los Gatos Tomato Huron 1991 512 84 tion operation at Fremont was also low (Lindeman, 2007). A
Products production line capable of producing 300 bottles of ketchup per
Ingomar Packing Los Banos 2000 410 67
hour was only producing 174 bottles. To raise the 58% efciency
Los Banos 1983 329 54
J.G. Boswell Bakerseld 2000 405 66 Heinz needed to increase the production volume. So it closed the
Corcoran 2008 333 55 2g.Tracy, California factory and moved that ketchup production
Average 398 65 volume to Fremont. In subsequent years it closed more factories
Food company Campbell Soup Dixon 1975 302 49 and moved more ketchup production volume to Fremont. Ten
Stockton 1967 171 28
Hunt Foods Oakdale 1970 206 34
years later, in 2007, the Fremont factory efciency exceeded 90%
Helm 1990 206 34 and the factory produced 80% of all the Heinz ketchup produced in
Unilever Stockton 1970 165 27 the United States. Fremont became the largest ketchup factory in
Del Monte Hanford 1976 88 14 the world. Because there was no local sourcing and no tomato
Average 190 31
processing operation, the factory imported 400 cases per day of
tomato paste (each case contained 3000 pounds of paste) from
California, produced from tomatoes grown there.
In addition to increasing efciency of ketchup production at
Fremont and, therefore, lowering cost, Heinz also needed to
increase efciency in the tomato processing operation in Califor-
nia. It did this by sourcing all tomatoes from California and
locating the tomato processing operation there. But instead of
using its own factories, Heinz closed its California processing
factories and contracted with private companies, called co-pro-
cessors, to process tomatoes into tomato paste using new tomato
Food company paste production technology.

4.3.3. Task 3: utilize new tomato paste production technology

Fig. 7. Tomato paste processing companies from 1997 to 2005 (Morning Star Packing, 2014).
acre; this increased to 40 t per acre in 2005, often as a result of better
farming practices developed by Heinz. But productivity increases could
not keep pace with cost increases. One example is the cost of seeds. This
cost tripled from $300 per packet in 2005 to $900 in 2012 (a packet
contains 100,000 seeds and about 50,000 seeds are planted per acre).
Heinz requires all its growers to purchase seeds from its subsidiary, the
HeinzSeed Company. The Campbell Soup Company is the only other
food company that breeds its own seeds and requires growers to
purchase its seeds. The price of tomatoes peaked at $80 in 2009 but
by 2012 it was down to $75. Table 7 shows that the price of tomatoes is
always lowest and productivity is always highest in California. This is
why Heinz now sources all its tomatoes from California.
4.3.2. Task 2: reduce factory processing cost
The growing season in California is longer than it is in Ohio,
Michigan, Indiana and Ontario. In California there is an early
Tomato paste production technology changed in the mid-1990s
when it became technologically possible to package tomato paste
in large (e.g. 3000 pound) sterile bins, storable for 18 months. The
bins could be transported to factories anywhere in the world and
the paste used there to produce ketchup and other nished
products. At the same time companies emerged in California
specializing in the new technology (Table 8). They received
tomatoes from nearby growers, processed and packaged the
tomatoes into tomato paste, and sold the tomato paste to large
food companies, such as Heinz and Kraft, as well as smaller food
companies. In California it was difcult to build factories because
of the state's demanding water and environmental regulations, so
the tomato paste companies increased their capacity by continu-
ally improving their operations and running them at very volume
and efciency. Many of the companies have such large-scale
operations that the tomato paste they produced was a commodity
product sold at a very low price.
Morning Star Packing is the largest and most efcient producer
of tomato paste. It has two factories. The Los Banos factory
processed 661 t of tomatoes into 108 t of tomato paste per hour
in 2014. (Notice that it takes approximately 661C1086 pounds
of tomatoes to produce one pound of tomato paste.) The Williams
factory is newer and has a higher capacity. The average capacity
for the private companies is 587 t of tomatoes or 98 t of tomato
paste per hour. Table 8 shows that grower-owned factories have
276 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278

smaller capacities and the food company processing factories are Table 9
the smallest. Actual prices of tomatoes, transportation, factory processing, and tomato paste
(Morning Star Packing, 2014).
In 1993 Heinz closed its tomato processing operations in 1a.
Fremont and 4b.Muscatine, refocused these factories on producing Year Pricesa
nished products, and outsourced the tomato processing opera-
tion to the private California tomato paste companies. In 1997 $ per ton $ per pound of tomato paste Totalb
Heinz closed its tomato processing factory in 2g.Tracy, California,
Tomato Transportation Factory processing
and in 2011 it closed its factory in 2h.Stockton, California, and
moved that production to the California tomato paste companies. 1975 $233 $0.70 $0.12 $0.70 $1.52
This left Heinz with one tomato processing factory in California, 1985 $113 0.341 0.084 0.386 0.811
2b.Escalon, and one in Canada, 1g.Leamington. In 2014 it closed 1990 100 0.301 0.061 0.347 0.709
1995 84 0.252 0.039 0.234 0.525
the Leamington factory. The Escalon factory remained open
2000 70 0.210 0.033 0.141 0.383
because it produced specialized tomato paste which could not be 2005 62 0.187 0.0365 0.167 0.390
produced by the tomato paste companies. The plentiful supply of 2010 73 0.219 0.0315 0.122 0.372
low-cost, high-quality tomato paste made it impractical for Heinz 2013 75 0.225 0.035c 0.096 0.356
and most other food companies to run their own tomato proces- a
All prices are in equivalent 2013 dollars.
sing operations. Fig. 7 shows that between 1997 and 2006 food b
Total tomato paste price per pound tomato price transportation pri-
companies moved about two-thirds of their tomato processing ce factory processing price.
c
operations from their own factories to the private and grower- The cost per truck to transport tomatoes from growers' elds to the factory is:
owned California tomato paste companies. $0.035C6 $0.0058 per pound of tomatoes
2000 pounds per ton
15 t per
trailer
2 trailers $348 per tandem trailer truck.

4.4. Global function production network: 2005 to present

By 2005 the Heinz ketchup network was transforming from a
multidomestic to a global function network. The tomato growing
function was focused in California. Growers were closely controlled
and tightly coordinated through contracts, proprietary Heinz seeds,
and special Heinz farming programs. The tomato processing function
was focused in the private California tomato paste companies. The
ketchup production function was focused primarily in the 1a.Fremont
factory. Scale was the common characteristic in the three functions.
Scale was necessary to achieve economies of scale and to justify large
investments in technology and improvements. To achieve scale (i) all
North America tomatoes were sourced from California growers, (ii)
the California tomato paste companies operated at very high volumes,
and (iii) the Fremont factory produced 80% of all the Heinz ketchup
produced in the United States.
It is instructive estimate the minimum efcient scale, MES,
(Weiss, 1975) for the tomato processing function and the ketchup
production function. The average capacity of a private tomato
paste factory is 587 t of tomatoes or 98 t of tomato paste per hour
(recall Table 8). For a grower-owned factory the average capacity is
lower: 398 and 65. A reasonable estimate, therefore, of MES for the
tomato processing function is 400 t of tomatoes or 83 t of tomato
paste per hour or 400 t of tomatoes per hour
(3 month
harvest
30 days per month
24 h per day)E one million tons
of tomatoes per harvest-year. Seven of the 16 factories in Table 8
operate at or above 400 t of tomatoes per hour and, therefore, are
able to give the lowest possible cost. California grew 12.6 million
tons of tomatoes in the 2014 harvest-year. If all of this is processed
into tomato paste then the required number of MES factories is
12.6C1 E13 factories. There are more than 13 factories in Cali-
fornia so there is unused capacity in the industry.
With respect to MES for the ketchup production function, the
Fremont factory uses 400 cases of tomato paste per day or 400
cases
3000 pounds per case
250 production days per year300
million pounds of tomato paste per year. So MES for the ketchup
production function is 300 million pounds of tomato paste per year
or 300
6 1.8 billion pounds of tomatoes per year. One pound of
tomato paste produces 1 pounds of ketchup, one pound of ketchup
produces 13 uid ounces of ketchup, and a mid-size ketchup bottle
is 32 ounces. So Fremont produces 300
1400 million pounds of
ketchup or 400
13C32 167 million (32-ounce) ketchup bottles
per year. If one ketchup bottling line produces 1000 bottles per hour,
then the Fremont factory needs 167 millionC(1000
24 h per
day
250 production days per year) 28 bottling lines.
A California MES tomato processing factory processes one
million tons of tomatoes per harvest-year. Fremont uses 1.8 billion
pounds 0.9 million tons of tomatoes per year. So Fremont
requires 0.9C1 90% of the capacity of one California MES tomato
processing factory. This does not mean it economical to locate a
tomato processing operation at Fremont. If Fremont has its own
tomato processing operation supplied by tomatoes from local
growers then it must process these tomatoes during a one month
harvest compared to the 3 month harvest in California. So a
Fremont tomato processing operation would require 0.9
3 3.1
California MES tomato processing factories. This would be a
prohibitively large and expensive facility.
Table 9 is a remarkable table from Morning Star Packing
showing the price of tomatoes, the price to transport tomatoes
from growers' farms to the Morning Star factory, and the factory
price to process tomatoes into paste. All prices are in equivalent
2013 dollars. For example, in 2013 the price paid to growers was
$75 per ton or $0.0375 per pound of tomatoes. Six pounds of
tomatoes are needed to produce one pound of paste, so the cost of
tomatoes is 6
$0.0375 $0.225 per pound of tomato paste. The
transportation price was $0.035 per pound of paste. The factory
cost including prot was $0.096 per pound; this is the cost to
process tomatoes into paste and package the paste into sterile
bins. The total price charged to Heinz was $0.225 $0.035
$0.096 $0.356 per pound of tomato paste.
Notice that $0.225C$0.356 63% of the price of tomato paste is
the amount paid to growers, and $0.035C$0.35610% is the amount
paid for transportation. This leaves only 27% for processing in the
factory. Clearly the factory must operate at a very high volume and
efciency to be protable. The 63% paid to growers in 2013 compares
to 0.70C1.5246% in 1975 and 54% in 2000; the percent paid to
growers has increased steadily. The percent paid to the factory was
0.70C1.52 46% in 1975, 37% in 2000, and 27% in 2013; the percent
paid to the factory has decreased steadily. Table 9 also shows that all
three costs: tomato, transportation, and factory processing, have
decreased signicantly. This means: (i) Heinz pays less for tomato
paste and, therefore, growers, transportation companies, and proces-
sing factories receive less, (ii) processing factories are squeezed
between growers who demand a higher percent of the price and
customers like Heinz who demand a lower price. Customers receive
the lower price because there is extra capacity in the industry and
because low-cost, tomato paste is now available from China. Heinz
already imports tomato paste from the Xinjiang Tunhe Company in
China (Wall Street Journal, 2007).
 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278
5. Strategic implications
Following the GPN framework we now make some strategic
observations on value, governance and ownership, power, and
embeddedness. Value is created from the tomatoes function (done
by California growers), the tomato paste function (done by
California tomato paste companies), and the ketchup production
function (done in the Fremont factory). Heinz governs the rst two
without owning either, and it governs and owns the last function.
Heinz governs the growing function by means of grower contracts,
proprietary seeds, and required farm practices. It governs the
tomato paste function by means of contracts, being a very large
customer, and sourcing from several processors and from China.
Heinz does not own these functions because it knows from past
experiences in Ohio and Ontario that when it owns factories then
worker unions and growers' associations publicize disputes which
adversely impacts retail sales of all Heinz products and, therefore,
forces Heinz to make concessions on wages and prices. Governing
without owning and dispersing functions in a global function
network is the best way to keep costs low.
Although Heinz prefers to use co-processors to produce tomato
paste this may change. (The appropriateness of using co-
processors in a global function network can be explored using
the framework in Rebolledo and Jobin, 2013.) The private tomato
paste companies are being squeezed between the growers and the
food companies. As their prot margins shrink it will be more
difcult to maintain high levels of quality. This could affect the
tomato paste Heinz receives and uses in its ketchup (and other
products). Other food companies (e.g. Campbell, Hunt) own large
tomato paste factories to preserve the high quality of their
products. Heinz has only one tomato paste operation, 2b.Escalon,
but it is small and specialized. It is likely that after Heinz upgrades
the ketchup production networks in its other business units (i.e.
Europe, Asia/Pacic, Rest of World) and gains more experience
with importing tomato paste from China, then it will acquire a
tomato paste company in California or China.
The corporate power of Berkshire and 3G at head-ofce and
collective global power of the large retail store chains (e.g. Wal-
Mart) will keep the pressure to reduce cost and to globalize very
high. This means the current global function production network
is a good network, but upgrading it to a global mixed network is
the strategy for the future. Integrating tomato paste imports from
China into its current North American operations is a sign that
Heinz has already started this upgrade. Heinz is also likely to
integrate its ketchup production network in North America with
its ketchup production networks in its business units in other
parts of the world.
Beginning in the early years of the company through to the
1990s Heinz became embedded in the three North America tomato
growing regions: California, the Ohio area, and Ontario. It is
interesting that Heinz has not been able to fully break out of this
embeddedness. When Heinz stopped sourcing tomatoes from the
Ohio area it did not close any of its four factories there. When it
stopped sourcing tomatoes from Ontario it did not entirely close
its Leamington factory; instead it sold the factory to a new
company (Highbury Canco) that immediately became a Heinz
co-processor (CBC News Windsor, 2014).
6. Conclusion and further research
The global production network framework and the manufac-
turing strategy framework greatly facilitated the systematic
description and analysis of the Heinz ketchup production network
over the time period from 1960 to 2015. The frameworks reveal
how and why Heinz built and changed its network, how it
277
currently manages its network, and how it is likely to change
and manage its network in the future. The analysis in this paper
illustrates how the frameworks can be used to analyze the past,
current, and future production networks of any large company.
This is particularly useful for agents (i.e. suppliers, employees,
customers, service providers, and infrastructure providers such as
governments) who have signicant interactions with these net-
works. The analysis in this paper is for a leading company in the
food industry. Other large companies in the food industry can be
studied to determine similarities and differences, and large com-
panies can be studied in other industries. The production net-
works of small companies can also be studied to determine the
effect of company size on the design and development of produc-
tion networks.
The Heinz ketchup network is an excellent global function produc-
tion network. This network-type is the reason Heinz is able to produce
the industry's leading product and be the industry's lowest-cost
producer. It would be interesting to compare this network to the
networks Heinz uses in its other business units (Europe, Asia/Pacic,
Rest of World) and for its other product families. It is likely that
different product families require different production network types
because of different food growing requirements, processing technology
constraints, competitive pressures, and so on. It would be interesting to
learn what production network types are favored by large food
companies for different food products. It would also be interesting
for researchers to examine what facility-types and capabilities are
favored by large food companies in various production network types.
For various reasons including the difculty of obtaining
detailed information, companies in the food industry are less
studied compared to companies in other industries (e.g. automo-
tive, aviation, electronics, machine tools). Given the importance of
the food industry and its presence in some form in every region of
the world, studies of the facility types, production network types,
competitive situations, and so on in companies in the food
industry is also an important area for future research.
References
Afonso, F., Fleury, M., 2009. Understanding the strategies of late-movers in
International Manufacturing. Int. J. Prod. Econ. 122, 340350.
Avella, L., Fernandez, E., Vazquez, C., 2001. Analysis of manufacturing strategy as an
explanatory factor of competitiveness in the large Spanish industrial rm. Int. J.
Prod. Econ. 72, 139157.
Cagliano, R., Acur, N., Boer, H., 2005. Patterns of change in manufacturing strategy
congurations. Int. J. Oper. Prod. Manag. 25 (7), 701718.
Cheng, Y., Farooq, S., Johansen, J., 2011. Manufacturing network evolution: a
manufacturing plant perspective. Int. J. Oper. Prod. Manag. 31 (12), 13111331.
Coe, N.M., Dicken, P., Hess, M., 2008. Introduction: global production networks
debates and challenges. Journal of Economic Geography 8, 267269.
Colotla, I., Shi, Y., Gregory, M., 2003. Operation and performance of international
manufacturing networks. Int. J. Oper. Prod. Manag. 23 (10), 11841206.
Demeter, K., 2003. Manufacturing strategy and competitiveness. Int. J. Oper. Prod.
Manag. 81, 205213.
Dicken, P., 2003. Global Shift: Transforming the World Economy, Fourth edition.
Sage, London.
Feldmann, A., Olhager, J., 2013. Plant roles: site competency bundles and their
relationships with site location factors and performance. Int. J. Oper. Prod.
Manag. 33 (6), 722744.
Ferdows, K., 1997. Making the Most of Foreign Factories. Harvard Business Review,
pp. 7388.
Friedli, T., Mundt, A., Thomas, S., 2014. Strategic Management of Global Manufac-
turing Networks: Aligning Strategy, Conguration, and Coordination. Springer,
Heidelberg, New York.
Garrido-Vega, P., Ortega Jimenez, C., Perez de los Rios, J., 2015. Implementation of
technology and production strategy practices: relationship levels in different
industries. Int. J. Prod. Econ. 161, 201216.
Hammamia, R., Frein, Y., Hadj-Alouane, A., 2008. Supply chain design in the
delocalization context: relevant features and new modeling tendencies. Int. J.
Prod. Econ. 113, 641656.
Berkshire Hathaway Annual Report, 2014. http://www.berkshirehathaway.com/
2014ar/2014ar.pdf, downloaded in February 2015.
Heinz Annual Report, 2009. http://www.heinz.com/data/pdf_les/BPI35859_
WO26_HEINZ-ANNUAL-VW.pdf, downloaded in February 2015.
278 J. Miltenburg / Int. J. Production Economics 168 (2015) 267278
Heinz Annual Report, 2012. http://www.heinz.com/data/pdf/Heinz_2012_Annual_
Report.pdf, downloaded in February 2015.
Henderson, J., Dicken, P., Hess, M., Coe, N., Yeung, H., 2002. Global production
networks and the analysis of economic development. Rev. Int. Polit. Econ. 9,
436464.
Heinz Security Exchange Commission form 10-K, http://www.heinz.com/our-com
pany/investor-relations/shareholder-information/sec-lings.aspx, downloaded
in February 2015.
Jaehne, D., Riedel, R., Mueller, E., 2009. Conguring and operating global production
networks. Int. J. Prod. Res. 47 (8), 20132030.
Lindeman, T.F., Fremont, Ohio: A visit to the heart of Heinz, Pittsburgh Post-Gazette,
April 17, 2007, http://www.post-gazette.com/business/businessnews/2007/04/
17/Fremont-Ohio-A-visit-to-the-heart-of-Heinz/stories/200704170329, down-
loaded February 2015.
Maritan, C., Brush, T., Karnani, A., 2004. Plant roles and decision autonomy in
multinational plant networks. J. Oper. Manag. 22 (5), 489503.
Miltenburg, J., 2005. Manufacturing Strategy, Second edition. CRC Press, London.
Morning Star Packing, 2014. Tomato paste and Processed Tomatoes Statistics, www.
morningstarco.com, downloaded in February 2015.
Muehlfeld, K., Weitzel, U., van Witteloostuijh, A., 2011. Mergers and acquisitions in
the global food processing industry in 19862006. Food Policy 36, 466479.
New York Times, Heinz reaches accord with migrant workers, April 12, 1987,
http://www.nytimes.com/1987/04/12/us/heinz-reaches-accord-with-migrant-
workers.html, downloaded February 2015.
Nichols, J., FLOC, Heinz agree to farm worker contract, Toledo Blade, Toledo, Ohio, p.
23, April 9, 1987, https://news.google.com/newspapers?nid=1350&dat=19870-
409&id=9FBPAAAAIBAJ&sjid=8AIEAAAAIBAJ&pg=6399,2718164&hl=en,
downloaded in February 2015.
Tomato News Online, About the tomato processing industry: updated 7th Novem-
ber 2014, http://www.tomatonews.com/resources.html, downloaded in Feb-
ruary 2015.
Porter, M., 1985. Competitive Advantage. The Free Press, New York.
Raine, G. Heinz hopes sweeter ketchup tomatoes reduce need for corn syrup, San
Francisco Chronicle, Wednesday October 3, 2007, http://www.sfgate.com/
business/article/Heinz-hopes-sweeter-ketchup-tomatoes-reduce-
need-2519820.php, downloaded February 2015.
Rebolledo, C., Jobin, M., 2013. Manufacturing and supply alignment: are different
manufacturing strategies linked to different purchasing practices? Int. J. Prod.
Econ. 146, 219226.
Reimer, S., 2007. Geographies of production. Prog. Hum. Geogr. 31 (2), 245255.
Rudberg, M., West, B., 2008. Global operations strategy: coordinating manufactur-
ing networks. 36. Omega: Int. J. Manag. Sci., pp. 91106.
Shi, Y., Gregory, M., 1998. International manufacturing networks to develop global
competitive capabilities. J. Oper. Manag. 16, 195214.
Statistics Canada. 1993 to 2014 Annual area, production and value of farm
vegetables, http://www5.statcan.gc.ca/subject-sujet/theme-theme.action?
pid=920&lang=eng&more=0&HPA, downloaded in February 2015.
Sydow, J., Frenkel, S., 2013. Labor, risk, and uncertainty in global supply networks
exploratory insights. J. Bus. Logist. 34 (3), 236247.
United States Department of Agriculture, National Agricultural Statistics Service,
August 2010, California Processing Tomato report, http://www.nass.usda.
gov/Statistics_by_State/California/Publications/Vegetables/201501ptom.pdf,
downloaded in February 2015.
Vereecke, A., vanDierdonck, R., 2002. The strategic role of the plant: testing
Ferdow's model. Int. J. Oper. Prod. Manag. 22 (5), 492514.
Vokurka, R., Davis, R., 2004. Manufacturing strategic facility types. Ind. Manag. Data
Syst. 104 (6), 490504.
Wall Street Journal. Tunhe may produce juicy gains as Chinese tomato exports grow,
December 28, 2007, http://www.wsj.com/articles/SB119878549382753423, down-
loaded in February 2015.
Weir, K., Kochhar, A., Lebeau, S., Edgeley, D., 2000. An empirical study of the
alignment between manufacturing and marketing strategies. Long Range Plan.
33 (6), 831848.
Weiss, L.W., 1975. Optimal plant size and the extent of suboptimal capacity. In:
Mason, R.T., Qualls, P.D. (Eds.), Essays on Industrial Organization. Ballinger,
Cambridge, MA.
CBC News Windsor. Michael Mullen (Heinz vice president) in Heinz sells
Leamington plant to Highbury Canco, May 21, 2014, http://www.cbc.ca/
news/canada/windsor/heinz-sells-leamington-plant-to-highbury-canco-250-
jobs-safe-1.2649162, downloaded in February 2015.
World Processing Tomato Council, World production estimate as of 30 October
2014, http://www.wptc.to/releases-wptc.php, downloaded in February 2015.