3. Structural changes in the seed industry
The seed industry has witnessed several phases of consolidation in the past three decades. This chapter presents profiles of the main firms in the global seed industry, with particular attention to recent changes. The drivers behind the observed pattern of change and consolidation are discussed in-depth and illustrated with a case study of the US cotton seed industry. Implications for the current merger wave are also examined.
3.1. The organisation of the seed industry
To understand structural changes in seed markets, it is useful to distinguish four stages in the seed supply chain (Fernandez-Cornejo, 2004[1]):
● Plant breeding, where R&D leads to new, improved varieties. This sector is highly concentrated. An even smaller set of firms is active in the development of GM traits.
● Seed production, which is typically outsourced by plant breeding firms to contract farmers.
● Seed conditioning, which is the stage where seeds are dried, cleaned, sorted, treated with insecticides and fungicides, and packaged for distribution and sale. Like seed production, this stage contains many smaller firms, often linked by contracts to large seed firms.
● Seed distribution, where seeds are sold to end users. Wholesale distribution is often controlled directly by large seed firms, sometimes through licensing agreements. At the local level, retail distribution of seeds often takes place through local intermediaries such as farmer-dealers or agricultural supply stores. The precise structure of the distribution stage differs from region to region. In some countries, agricultural cooperatives play an important role in distribution of agricultural inputs, including seeds (Syngenta, 2016[15]).
By way of illustration, the maize seed sector in France comprises 12 plant breeding enterprises, 40 seed production firms (which in turn contract with 4 615 farmers), and 4 842 sales outlets (Gnis, 2017[34]). For the European Union as a whole, an estimated 7 000 firms are active in the seed industry across the various stages of the supply chain. Poland, Romania and Hungary together account for 2 800 such firms, most of which are small and medium enterprises (Ragonnaud, 2013[17]).
Research and development can be undertaken by the public sector (e.g. at universities) as well as by private firms, and is typically costly and time-consuming. Traditional plant breeding in particular takes several years, as it most often requires repeated crossing and selection. On average, it takes more than ten years after the first crossing before any actual sales are made (KWS, 2017[35]). Recent breakthroughs using the so-called New Plant Breeding Techniques (NPBTs) could potentially lead to significant reductions in the time needed to develop new varieties (Schaart et al., 2015[11]) (Scheben and Edwards, 2017[33]). For research on genetically modified organisms, significant additional costs are incurred for safety assessments and other tests required by regulatory bodies (Kalaitzandonakes, Alston and Bradford, 2007[36]).
The category of plant breeders includes large, well-known companies such as Monsanto, as well as a range of smaller niche players. There are also many smaller seed companies that do not engage in R&D, but produce and sell seed under licensing or other commercial arrangements, and (at the other end of the spectrum) there are small- and medium-size agricultural biotechnology companies which engage in R&D but often have little or no sales yet (Heisey and Fuglie, 2011[7]).
3.2. Leading firms in the global seed industry
This section provides background on the leading firms in the industry. Firms involved in recent mergers and acquisitions are discussed first (in chronological order according to the date when the initial bid or proposal was announced); two other important players (Limagrain/Vilmorin and KWS) are then discussed.
DowDuPont
The merger of Dow and DuPont was initially announced in December 2015, and officially concluded on 1 September 2017. The new company, DowDuPont, consists of three main divisions which will eventually become separate and independent companies. Its agriculture division, under the name of Corteva Agriscience, corresponds to the activities of DuPont Pioneer, DuPont Crop Protection and Dow AgroSciences. Two other divisions focus on materials science and specialty products (DowDuPont, 2018[37]).1
Before merging, Dow’s agricultural sales were mostly centred on crop protection chemicals while DuPont’s agricultural sales were dominated by seeds (notably through its Pioneer brand). Figure 3.1 shows a pro forma representation of combined sales of Dow and DuPont based on 2016 sales data, not taking into account any divestitures. Before divestitures, the combined agricultural sales of Dow and DuPont were around USD 16 billion. As shown in Panel (a), Dow and DuPont have complementary profiles and the joint company has a more balanced segment split as a result. As shown in Panel (b), within seeds and traits, maize seeds are by far the most important source of revenues, with soybean seeds a distant second. Within crop protection, most revenues come from herbicides (where Dow is traditionally strongest) and insecticides (where both Dow and DuPont historically had important sales).
In terms of geographic split (Figure 3.2), DuPont’s agricultural sales before the merger were mostly concentrated in North America, which accounted for 54% of 2016 sales. North America was also the centre of gravity for Dow Agricultural Sciences (41% of 2016 sales), but compared to DuPont, a larger share of revenues came from Latin America.
Before taking into account the divestitures, the combined Dow and DuPont agricultural operation is therefore geographically concentrated in the Americas, with almost half of sales in North America and 22% in Latin America. Europe (including India for Dow) accounts for 19% and Asia Pacific for 10%.
The merger of Dow and DuPont was scrutinised by competition authorities. The European Commission conditionally approved the merger in March 2017 subject to significant divestitures. DuPont agreed to divest a large part of its crop protection business, including its global crop protection R&D organisation. In May 2017, Dow and DuPont received approval in Brazil conditional on divesting assets related to maize seeds in the Brazilian market. In June 2017, the United States Department of Justice announced the conclusions of its own investigation, and required divestitures similar to those required by the European Commission. Following the divestitures, total sales in DowDuPont’s agriculture division are estimated at around USD 14 billion (compared to around USD 16 billion in the pro forma calculation shown in the charts). Data presented in the 2017 Annual Report of DowDuPont do not allow for a more detailed analysis.
Note: Pro forma estimates based on the 2016 sales figures for the companies, not taking into account divestitures. Data is approximate and based on company materials.
Source: Company annual reports; Dow presentation at Bank of America Merrill Lynch Global Agriculture Conference, March 2017.
Note: Pro forma estimates based on the 2016 sales figures for the companies, not taking into account divestitures. Europe for Dow refers to Europe, Middle East, Africa and India. Data is approximate and based on highly aggregate numbers in company materials.
Source: Company annual reports; Dow presentation at Bank of America Merrill Lynch Global Agriculture Conference, March 2017.
ChemChina-Syngenta
After a failed bid by Monsanto on Syngenta in 2014-15, ChemChina (China National Chemical Corporation) launched a bid for Syngenta in February 2016, offering USD 43 billion for the company. Regulatory approval was obtained in April 2017 from EU and US regulators, but under the condition of divesting certain parts of the ChemChina and Syngenta pesticide businesses.
ChemChina is the largest chemical firm in the People’s Republic of China (hereafter “China”) and operates a wide range of businesses from basic chemicals to high-end manufacturing.2 Although it is a state-owned company, ChemChina has been described as “functioning as an aggressive private business” and a “state enterprise in name only” (Weinland and Hornby, 2017[38]). The Syngenta acquisition is the largest overseas acquisition by a Chinese firm to date. In June 2018, ChemChina announced a merger with Sinochem, another large state-owned chemical conglomerate (Reuters, 2018[39]). The merger will create the world’s largest industrial chemicals group. One reason for the merger is reportedly the need to ensure that ChemChina has sufficient financial strength to absorb Syngenta (Weinland and Hornby, 2017[38]).
Syngenta’s CEO, Erik Fyrwald, has described ChemChina’s acquisition of Syngenta as a strategic step to ensure food security for China through a two-pronged strategy. The goal, according to Fyrwald, is not only to acquire leading technology to improve productivity in China, but also to ensure continuous development of technology to improve productivity elsewhere in order to maintain potential sources of food imports for China (Colvin, 2017[40]) (Tsang, 2017[41]).3 Historically, ChemChina has invested little in agricultural R&D; before the Syngenta acquisition, its core activity in agriculture was in the generic pesticide market through its Adama subsidiary. This makes the ChemChina-Syngenta acquisition different from the DowDuPont and Bayer-Monsanto transactions, which combined two R&D-intensive firms.
Syngenta’s sales are mostly concentrated in Europe, the Middle East and Africa (EMEA), Latin America, and North America (Figure 3.3); Asia is currently less well-served. Seeds account for about 23% of Syngenta’s global revenues, but this share is smaller in Asia Pacific (15%). From Syngenta’s point of view, the acquisition by ChemChina could open opportunities for sales growth in Asia, in particular in the seeds segment.
Note: EMEA is Europe, Middle East and Africa.
Source: Company annual reports.
ChemChina’s subsidiary Adama sells generic pesticides in the European Union and the United States (among other markets). Both in the United States and the European Union, there was an overlap between Adama’s generic portfolio and Syngenta’s portfolio of pesticides; in the European Union there was additional concern around Adama’s plant growth regulator business. In the United States, competition authorities required ChemChina to divest three pesticide products. In the European Union, ChemChina has divested a significant part of the Adama pesticide business, as well as some Syngenta pesticides, several generic pesticides which were under development, and part of Adama’s plant growth regulator business (European Commission, 2017[42]) (Bartz, 2017[43]). As ChemChina has no seed sales, there was no competition conflict with Syngenta’s seed businesses.
Bayer-Monsanto
In September 2016, Bayer and Monsanto announced a merger agreement whereby Bayer would pay USD 57 billion to acquire Monsanto. Materials made available to investors describe the merger as bringing together “two different, but highly complementary businesses” and notes that “[t]he combined business will benefit from Monsanto’s leadership in Seeds & Traits and Climate Corporation platform along with Bayer’s broad Crop Protection product line across a comprehensive range of indications and crops in all key geographies” (Bayer and Monsanto, 2016[44]). To obtain regulatory approvals, Bayer had to divest most of its seeds and traits activities and important parts of its herbicide business, which altogether accounted for around USD 2.5 billion in sales. After BASF acquired these activities from Bayer, the merger of Bayer and Monsanto was officially completed in June 2018.
The Monsanto acquisition transforms Bayer from a firm mostly focused on pharmaceuticals and consumer health to a firm with a roughly equal share of sales in healthcare and in crop science and seeds. The new Bayer Crop Science, including Monsanto, has combined sales of around USD 23 billion (using pro forma 2017 sales after divestitures). This makes Bayer the largest firm in the industry.
Before the divestitures, Bayer’s CropScience division derived most of its revenues from agricultural chemicals (notably herbicides and fungicides), with a smaller contribution of seeds and GM traits (Figure 3.4). Monsanto’s profile was complementary, as it derived most of its revenues from the sales of seeds and GM traits and, to a lesser extent, herbicides. Monsanto had little or no revenues from fungicides, insecticides, or seed treatment.
In terms of geographic split, the firms also had different profiles. While the sales of Bayer CropScience were diversified across different regions, Monsanto’s sales were heavily concentrated within North America. This region accounted for about two-thirds of Monsanto’s revenues, with sales in the United States alone accounting for 56% of the firm’s revenues. Latin America was the second most important region, with 13% of global revenues coming from Brazil and 6% from Argentina. However, sales in Europe, the Middle East and Africa (EMEA) and Asia Pacific were considerably smaller than Bayer’s.
As a first approximation, Monsanto could be characterised as a company selling GM seeds and herbicides in the Americas, whereas Bayer could be characterised as a company selling mostly agricultural chemicals across a broad range of regions, but with relatively less presence in the United States. At this high level of aggregation, Bayer and Monsanto appear complementary in terms of product offering and regions, although overlap existed in several markets.
Note: Pro forma estimates based on the 2017 sales figures for the companies. The segment and region split of EUR 2.2 billion (USD 2.5 billion) in divested Bayer sales is approximate and based on BASF materials. Assumed segment split of the divested businesses: EUR 800 million in herbicides and EUR 1.4 billion in seeds and traits. Assumed split by regions: EUR 1.4 billion in North America; EUR 300 million in South America; EUR 300 million in EMEA; EUR 200 million in Asia Pacific. These assumptions are consistent with investor materials shared by BASF during a 27 July 2018 conference call with financial analysts. Financial data converted using an average exchange rate of USD/EUR 1.15. EMEA stands for Europe, Middle East and Africa. North America includes Mexico.
Source: OECD estimates based on company annual reports, Bayer (2016[45]) and BASF Q2 2018 Analyst Conference Call handout.
The divestiture of several Bayer businesses to BASF (Table 3.1) does not fundamentally change this picture. Bayer sold its non-selective herbicide (Liberty), the LibertyLink herbicide tolerance technology, nearly all of its global soybean and rapeseed/canola seed business, its cotton seed business (except in India and South Africa), its global vegetable seeds business, and several related R&D assets.4 An estimate of the corresponding sales divested by Bayer are shown in Figure 3.4. According to these estimates, the divested businesses correspond to nearly the entire seeds and traits business of Bayer and a considerable share of its herbicide business, and around half of Bayer’s sales in North America. Despite these significant divestitures, Bayer and Monsanto continue to have complementary profiles, both in terms of regions and in terms of products.
In the future, the firms expect to move beyond simply a combined offering of current products towards developing “integrated solutions” to include seeds and traits, crop protection, and digital farming (Bayer and Monsanto, 2016[46]). The combined firm would have a pro forma R&D budget of USD 2.8 billion (12% of sales) compared with estimated budgets of USD 1.8 billion for DowDuPont (11% of sales, not taking into account potential R&D reductions underway now) and USD 1.4 billion for ChemChina-Syngenta (8.5% of sales). The combined R&D capabilities would include more than 35 R&D sites and around 175 breeding stations.
BASF
BASF was traditionally the sixth member of the Big Six of seed and agrochemical suppliers. In contrast to the other players, which combined seeds and chemicals sales, BASF’s agricultural sales have historically focused almost exclusively on agricultural chemicals (Figure 3.5). Not including the recently acquired Bayer business, the 2017 sales of BASF’s Agricultural Solutions division amounted to USD 6.6 billion, with most of the revenues coming from fungicides and herbicides. Europe and North America were the main markets for its agricultural sales, accounting for more than two-thirds of revenues.
Note: Using an average exchange rate of USD/EUR 1.15. “S. Am & MEA” is South America, the Middle East, and Africa. Segment and region split of EUR 2.2 billion of acquired Bayer sales is approximate. Assumed segment split: EUR 800 million in herbicides and EUR 1.4 billion in seeds and traits. Assumed region split: EUR 1.4 billion in North America; EUR 300 million in South America, the Middle East and Africa; EUR 300 million in Europe; EUR 200 million in Asia Pacific. These assumptions are consistent with investor materials shared by BASF during a 27 July 2018 conference call with financial analysts.
Source: Company reports and BASF Q2 2018 Analyst Conference Call handout.
BASF did not participate in the earlier rounds of mergers and acquisitions that created integrated seeds-and-pesticides firms, preferring instead to focus on crop protection chemicals. The transfer of several Bayer businesses changes the picture. BASF not only acquires Bayer’s non-selective herbicide (Liberty), but also most of Bayer’s seeds and traits business. The seeds and traits segment has become a significant part of BASF’s agricultural sales, accounting for some 18% of sales (Figure 3.5). While this transaction increases BASF’s portfolio, the firm will nevertheless remain a small player compared to Bayer-Monsanto, DowDuPont and ChemChina-Syngenta.
Limagrain/Vilmorin
While less well known than the Big Six, the French cooperative Limagrain is a major international player in seeds due to its ownership of Vilmorin, a leading plant breeding firm initially founded in the 18th century by the botanist to the French King Louis XV (Kingsbury, 2009[30]). Total sales of Vilmorin were around USD 1.6 billion in 2017-18. In addition, Vilmorin owns 50% of AgReliant, a joint-venture with KWS (discussed in more detail below). Adding its share of AgReliant sales, total sales reach USD 1.9 billion. Field seeds account for 55% of this total, while vegetable seeds contribute 42%. The remainder is accounted for by sales of gardening seeds and other activities (Figure 3.6).
Reflecting its origins in France, 50% of total sales occur in Europe. Vilmorin also has considerable sales in the Americas (34%), notably through AgReliant, but the firm’s footprint in Africa, the Middle East, and Asia Pacific is limited (Figure 3.7). The geographic split for field seeds is different from that of vegetable seeds. While field seeds sales are heavily concentrated in Europe and the Americas (accounting for 90% of field seeds sales), the sales of vegetable seeds are more evenly spread out across regions. Europe and the Americas each contribute about one-third of total vegetable seeds sales, with the remainder split between Africa and Middle East, and Asia and Pacific.5
Within field seeds, two-thirds of sales are for cereals (maize, wheat and barley) and 15% for sunflower. AgReliant’s field seeds sales in North America (half of which are included in Vilmorin’s sales figures in this report) are heavily concentrated in maize (78%) and soybeans (21%).
Note: Using an average exchange rate of USD/EUR 1.15. AgReliant is a 50/50 joint venture with KWS.
Source: OECD estimates based on company annual report.
Note: Using an average exchange rate of USD/EUR 1.15. Data refer to 2017-18. Totals may differ due to rounding.
Source: OECD estimates based on company annual report. Data include 50% of sales of AgReliant (a 50/50 joint venture with KWS).
KWS
The German-based seed firm KWS, founded in 1856, focuses on three major segments: maize and oilseeds, sugar beet, and (non-maize) cereals (Figure 3.8). Nearly half of the USD 1.2 billion in KWS sales come from maize and oilseeds, while sugar beet accounts for 42% of sales. In addition, KWS is 50% owner of AgReliant (jointly owned with Vilmorin). Adding a proportionate share of AgReliant sales to KWS sales figures raises total sales to USD 1.6 billion, of which 59% are in maize and oilseeds.
Note: Using an average exchange rate of USD/EUR 1.11. Data refer to 2016-17. AgReliant is a 50/50 joint venture with Limagrain/Vilmorin.
Source: OECD estimates based on company annual report.
Germany is a major market for KWS, accounting for 21% of the firm’s sales. Other European countries represent 43% of the total, while the Americas account for 30%. After adding the KWS share of AgReliant sales, the importance of the Americas grows to 46%.
KWS has a remarkably strong position in sugar beet, where it estimates it has a 55% global market share. In the EU-28, KWS estimates its market share to be around 44%, while this share in North America is estimated to be well over 80% (KWS, 2017[35]).
In 2018, KWS attempted to purchase Bayer’s vegetable seeds business, which was being sold as part of Bayer’s preparation for its merger with Monsanto. This acquisition would have diversified KWS’s activities away from field crops and would have led to a significant increase in size for KWS (Burger and Weiss, 2018[47]). Its bid, however, was not successful.
Leading firms after the mergers
Figure 3.9 shows the new Big Six in the seeds industry after the mergers, using pro forma estimates of 2017 sales after taking into account the estimated impact of mergers and divestitures. Bayer-Monsanto is the largest player, with roughly equal shares of sales coming from seeds and biotech versus agricultural chemicals. ChemChina-Syngenta is second, at about a third smaller than Bayer-Monsanto, but mostly focused on agricultural chemicals. DowDuPont is the third major player, with a roughly equal split between seeds and biotech. Following the acquisition of Bayer assets, BASF has become the fourth player in the sector, although its total sales are less than half of those of Bayer-Monsanto. Finally, both Limagrain/Vilmorin and KWS, while important players, are small in comparison with the market leaders. Each firm has less than one-tenth the sales of Bayer-Monsanto.6
Note: Pro forma estimates based on the 2017 sales figures for the companies (or most recent data available). Bayer-Monsanto includes estimated effect of divestitures (about USD 2.5 billion). ChemChina-Syngenta includes an estimated USD 4.3 billion in agro-chemical sales of ChemChina. DowDuPont based on pro forma 2017 data reported by DowDuPont (accounting for divestitures). BASF includes estimated effect of acquired Bayer business. Data for Vilmorin and KWS include sales of AgReliant.
Source: OECD estimates based on company annual reports; Colvin (2017[40]) for ChemChina.
3.3. Drivers of structural change and consolidation
Schenkelaars et al. (2011[29]) describe the evolution of the global seed industry in three waves.
● A first wave occurred in the 1930s, when hybrid seed was introduced. New commercial seed firms emerged (including Pioneer Hi-Bred, now part of DowDuPont), which adapted and improved varieties developed through public research.
● A second wave occurred in the 1970s, around the same time as the intellectual property regime for plant breeding was strengthened via plant breeders rights (PBRs) and patents. Several pharmaceutical, petrochemical and agrochemical companies in the United States and Europe started a process of mergers and acquisitions. However, many seed firms remained independent and maintained their market position. Over time, several multinationals divested their seed assets.7
● A third wave of structural change started in the 1980s and was driven by biotechnology. A small number of large agrochemical multinationals invested heavily in these new technologies, both through in-house R&D and by acquiring smaller players. This third wave was characterised by strategic mergers and acquisitions to obtain access to varieties, traits, and tools, leading to the current constellation of integrated firms.
Figure 3.10 presents the “family tree” of the main players in the seed industry today, including recent mergers. An open question is whether the industry is witnessing a fourth wave of consolidation caused by data-driven precision agriculture.
Fulton and Giannakas (2001[48]) argue that structural changes in the seed and biotech industry are best understood as a mix of horizontal and non-horizontal combinations, which each have their own set of drivers. A horizontal merger combines firms active in the same or similar markets, and aims to achieve economies of scale (when costs can be spread over a larger production volume of the same product) and economies of scope (when costs can be spread over different types of products). In the seed and biotech industry these economies of scale and scope exist according to Fulton and Giannakas (2001[48]) because of the presence of large sunk costs related to R&D and regulatory costs of GM. In contrast, non-horizontal mergers combine firms active in different markets – for example, between a supplier and a customer (a vertical merger) or between firms producing complementary products. In the seed and biotech industry, non-horizontal mergers and acquisitions are driven by various complementarities.
The family tree of modern seed firms shows both horizontal and non-horizontal combinations. Although each of the major players has followed a distinct trajectory, paths typically show a clear link between plant breeding, biotech, chemicals, and pharmaceuticals. Syngenta, for instance, can be traced back to a chemical company (Imperial Chemical Industries), a pharmaceutical company (Novartis), and a seed company (Vanderhave). At the same time, Syngenta’s 2004 acquisition of Adventa/Garst can be seen as a horizontal move to expand the scale and scope of its seed business.
Note: Only major acquisitions in seeds, agrochemicals and biotech are shown.
Source: OECD based on Fernandez-Cornejo (2004[1]), Heisey and Fuglie (2011[7]), Howard (2009[49]), Crunchbase (2017[50]), Wesseler et al. (2015[26]), company websites.
These drivers of consolidation are also cited by industry participants. Schenkelaars et al. (2011[29]) asked executives of nine seed firms to rank the different causes of recent industry consolidation. Their seed firms included Monsanto, DuPont-Pioneer, Syngenta, KWS, and Limagrain, as well as smaller firms such as Rijk Zwaan (which specialises in vegetable seed). Most respondents agreed that an increase in plant breeding R&D costs was a key driver (Figure 3.11). This was true even for vegetable seed firms where GM was not relevant at the time of the survey. In addition to continuing efforts in GM, other seed firms were investing in advanced non-GM technologies. Other causes which were ranked as relevant include the high costs of applying GM technology, the extension of the patent system to cover plants (see Box 3.1), and regulatory requirements for GM technology. One firm mentioned the cost of access to genetics (finished lines) and germplasm for breeding uses, as well as legal costs as drivers of industry consolidation.
In terms of the distinction made by Fulton and Giannakas (2001[48]), the increasing costs for plant breeding and GM technology, and costs related to regulatory requirements for GMOs are drivers of horizontal mergers and acquisitions. Patent rights, access to genetics and germplasm, and the associated legal costs of navigating intellectual property claims can be seen as drivers of non-horizontal mergers and acquisitions (although they may play a role in horizontal mergers and acquisitions).
Note: Based on interviews with executives of seed firms (DuPont-Pioneer, Bayer-Nunhems, Illinois Foundation Seed, KWS, Limagrain, Monsanto, Rasi Seeds, Syngenta, Rijk Zwaan). Possible causes ranked from 1 (low) to 5 (high). One firm added Costs of access to genetics (finished lines) and germplasm for breeding uses and Legal costs as drivers, both with a score of 4 (not included in the above chart).
Source: Schenkelaars et al. (2011[29])
Sunk fixed costs as driver of horizontal mergers and acquisitions
The existence of sunk fixed costs implies that firms can lower their average costs by increasing the scale or the scope of their operations.88 Two types of sunk fixed costs often cited in the context of the seed industry are regulatory costs and R&D costs.
Regulatory costs for GM technology
Regulatory costs associated with the approval of a genetically modified organism are high both in the United States and the European Union. A detailed analysis of the duration of the regulatory process by Smart et al. (2017[51]) shows the process takes an average of almost 2 500 days in the United States and 1 800 days in the European Union. These different durations are not directly comparable, but they do show that the approval process is long.99
The financial costs of compliance have been estimated at between USD 6 and USD 15 million for genetically modified maize traits (Kalaitzandonakes, Alston and Bradford, 2007[36]). Industry-commissioned studies have put the regulatory cost as high as USD 35 million, or 26% of the total cost of introducing a new GM crop (Phillips McDougall, 2011[52]). Some industry executives interviewed by Schenkelaars et al. (2011[29]) placed regulatory costs at USD 100 million, while others mentioned USD 15 to USD 30 million. Some caution is required in interpreting these numbers as it can be difficult to allocate the costs of various activities and processes to “regulatory” as opposed to “non-regulatory” factors. Moreover, costs for a single-country approval will differ from those needed for global regulatory approval. Despite difficulties in measuring regulatory costs precisely, there is little doubt that compliance requires considerable investment, which adds to the sunk fixed costs of developing and introducing a GM trait. It has been argued that high regulatory costs discourage investment in genetic modification for crops with small markets; this phenomenon has been documented for specialty crops (Miller and Bradford, 2010[53]).1010
Regulatory costs could contribute to horizontal mergers and acquisitions through several channels. First, larger firms may be more efficient in dealing with regulatory procedures, for instance because larger firms can employ full-time specialists. Second, a larger firm can expect greater sales for a given product introduction, all else equal, which means a given regulatory cost burden can be spread over a greater sales volume. Third, mergers could allow firms to reduce the rate of innovation, thereby avoiding the sunk costs of product innovations, including regulatory costs.
It is doubtful, however, whether regulatory costs by themselves can explain ongoing consolidation, and in particular whether they can explain the current mergers. Even without regulatory costs, firms would face high R&D costs to develop and introduce GM traits. Moreover, over time the total market for genetically modified seeds has grown strongly, as documented above. In theory, this growth in market size should enable more firms to recover the regulatory costs associated with GM technology, all else equal. Regulatory costs may act as a barrier to entry and may reduce innovation in products with small markets, but it is not clear whether they are an important driver of current consolidation. Put differently, regulatory costs may contribute to a higher level of market concentration than would otherwise be the case, but they do not necessarily explain increases in the level of market concentration in recent years.
Increasing R&D costs
By definition, a fixed cost is a cost incurred by firms independent of how much they produce. This does not mean that all fixed costs are outside of the firm’s control. Some fixed costs (e.g. equipment and infrastructure) may be related to the minimum efficient scale of operations, leaving the firm with little choice on how much to spend on these items. But other fixed costs depend on strategic decisions of firms, as is the case for advertising budgets or R&D efforts. Such endogenous fixed costs have different implications for market structure compared with exogenous fixed costs (Sutton, 2007[54]).
If a fixed cost is exogenously given to the firm, then an increase in market size makes it possible for new firms to enter the market. Hence, a growing market would then be associated with an increase in the number of firms or, equivalently, a decrease in market concentration.
In a setting with endogenous fixed costs, firms use their investments in advertising or R&D as strategic tools to increase consumers’ willingness to pay, improve the quality and range of their product offering, and/or reduce their costs. When the size of the market increases, so does the potential payoff to the firm of capturing a greater market share through higher advertising spending or by offering an improved product. A greater market size is then associated with increased spending on endogenous fixed costs by existing firms. This creates a barrier to entry for new firms, and consequently the number of firms in the industry does not increase when the market grows. In fact, as the market grows and firms escalate their spending on endogenous fixed costs, market concentration may increase.
Anderson and Sheldon (2017[55]), analysing the US market for GM maize seed, find empirical evidence that these markets are indeed characterised by endogenous fixed investments in R&D. They conclude that the concentration of R&D activity and of sales are probably driven by endogenous investments to introduce better GM traits. In this case, a certain level of concentration is to be expected as a by-product of the same dynamic that ensures high R&D spending on improved GM traits. The increasing market concentration could come about as some firms are driven out of the market and/or as some firms acquire or merge with competitors to reduce the burden of R&D spending. Hence, the observed increase in market concentration in the seed market could indeed be consistent with endogenous R&D spending – an explanation which matches the views of industry participants as reported in Figure 3.11.
Complementarities as a driver of non-horizontal mergers and acquisitions
The seed industry has seen considerable non-horizontal activity, where seed, biotech, and chemical firms combine in various ways. The global seed industry is characterised by several complementarities which help to explain these non-horizontal mergers and acquisitions in the past decades (Heisey and Fuglie, 2011[7]). These complementarities are also important in evaluating the potential effects of mergers, and can be grouped into two general sets.
Tools, traits and germplasm
A first set of complementarities arises from the process of creating genetically modified seeds. The development of modern biotechnology created a new market for genes conferring valuable traits, and for platform technologies or research tools to create such traits. However, valuable traits need to be inserted in high-quality varieties to be useful to farmers. As a result, there is a strong complementarity between tools, traits, and varieties (germplasm) (Heisey and Fuglie, 2011[7]).
Graff, Rausser and Small (2003[56]) demonstrated empirically that mergers and acquisitions in the seed and biotech industry in the 1990s reflect complementarities between intellectual property assets in tools, traits and germplasm. Using patent data, they show that firms seek diversified portfolios (combinations of tools, traits and germplasm) through both in-house R&D as well as through mergers and acquisitions to achieve coordination between complementary intellectual assets.
Seeds and chemicals
A second set of complementarities is between seeds and chemicals. There are several explanations for this link. First, there is the possibility of creating complementary products, such as herbicide-tolerant seeds and herbicides. Secondly, there are potential economies of scope in marketing. A third reason is that chemical companies invested in seed as a defensive move when it became clear that genetically modified seeds could compete with crop protection chemicals (Heisey and Fuglie, 2011[7]).
The current consolidation wave demonstrates these complementarities. Monsanto, which has a relatively stronger position in seeds and biotech, is combining with Bayer, which has a stronger position in agricultural chemicals. Likewise, DuPont’s strong position in seeds and biotech is now combined with Dow Chemical’s relatively stronger position in agricultural chemicals. ChemChina was only active in agricultural chemicals but has acquired Syngenta which has a strong position in seeds and biotech. Finally, the seeds and biotech business divested by Bayer were acquired by BASF, which was previously active in agricultural chemicals only.
A further example is found in the historical evolution of Monsanto. Founded in 1901 as a chemical company, Monsanto focused on chemicals and pharmaceuticals for most of its history. In the early 1980s, it started to invest in genetic modification. The need to combine traits with germplasm and a distribution network led to an ambitious acquisition strategy in the 1990s to build Monsanto into a seed company. In the span of only two years (1996-1998), Monsanto acquired a range of seed companies that included the international seed businesses of DeKalb and Cargill, building a strong position in seed markets while continuing to invest in the acquisition of other biotechnology firms (Figure 3.10). In 1996, genetic modification by Monsanto led to the introduction of Roundup Ready soybeans, resistant to Monsanto’s glyphosate-based herbicide Roundup, first introduced in 1974.
The role of potential complementarities (and substitution effects) between biotechnology and agricultural chemicals was first highlighted by Just and Hueth (1993[57]). Certain developments in biotechnology are complementary with sales of agricultural chemicals, as in the case of herbicide-tolerant seeds and herbicides. In contrast, other developments in biotechnology are substitutes for agricultural chemicals, as in the case of insect-resistant seeds which are a substitute for insecticides.
Firms selling agricultural chemicals will tend to invest in biotechnology complementary to their existing products. As the integrated firm internalises the spillover effects of the complementary products, it can generate greater chemical sales by selling more biotechnology and vice versa. For such products, mergers between chemical and biotech firms would increase investment in R&D and total output, which would have a positive effect on social welfare.
If, however, a firm sells chemicals and biotechnology which are substitutes (e.g. traditional insecticides and insect-resistant seed), it may try to keep the supply of biotechnology low to avoid “cannibalising” its sales of chemicals. In such a scenario, the potential welfare effects of a merger between a chemicals firm and a biotech firm are less clear; the merger could be an attempt to reduce competitive pressure. The relative importance of these complementarity and substitution effects is therefore essential in order to evaluate the welfare effect of a merger.
Other possible complementarities
In the 1990s, industry observers speculated there might be complementarities between biotechnology and pharmaceuticals, which could give rise to integrated life science companies. However, this trend did not materialize. Instead, firms such as Dow, BASF and DuPont divested their pharmaceutical activities (Heisey and Fuglie, 2011[7]). Again, Monsanto is a representative example. The firm, which previously had some pharmaceutical activity, merged in 1999 with Pharmacia & Upjohn to create the chemical-pharmaceutical giant Pharmacia. By 2002, however, the agricultural business was spun off as the “new” Monsanto, severing the link with the pharmaceutical side of the business (Figure 3.10).1111
A potential new type of complementarity may be emerging in the form of “digital farming”, a technological trend to use big data techniques to allow precision farming (Kempenaar et al., 2016[58]). The promise of digital farming is that it could combine detailed data from thousands or millions of individual farmers, together with detailed weather data to uncover through algorithms which seeds, agrochemicals or other agronomic practices are optimal in which setting. This could also deliver precise advice to farmers, tailored to the specific characteristics of their plots and crops.1212
While many smaller firms appear to be active in this field, there is a clear trend for large players in agricultural input industries to enter the market (Philpott, 2016[59]). For instance, Syngenta offers digital farming solutions through its AgriEdge Excelsior product. DuPont expanded its own precision farming offering in 2014 through an agreement with the weather and market data provider DTN, while in 2016 Dow AgroSciences launched a “precision agronomy program” based on big data.
Monsanto in particular has invested heavily in precision agriculture, first through the acquisition of Precision Planting in 2012 and then through the USD 930 million acquisition of The Climate Corporation, a Silicon Valley firm specializing in weather prediction. It also entered into an agreement with the leading agricultural equipment manufacturer John Deere in 2015 whereby John Deere would acquire the equipment business of Precision Planting from Monsanto, while providing Monsanto with real-time data from certain John Deere farm equipment. However, this agreement was challenged by the US Department of Justice and subsequently terminated in May 2017.
Monsanto’s leading position in digital farming is cited by Bayer as an important reason for its bid. To some industry observers, the Bayer-Monsanto merger indicates that digital farming may be a transformative technology similar in its impact to GM technology two decades ago (Gullickson, 2016[60]).
Complementarities between digital farming and the seed and agro-chemical industry can arise from several mechanisms. First, digital farming solutions are sold to the same group of customers as seed and agro-chemicals, creating possibilities for joint marketing. Second, data from thousands of farmers can provide an unprecedented dataset of the performance of crops as a function of soils, weather conditions, farming practices, and types of seeds and agro-chemicals used. This could allow for more tailored marketing efforts (e.g. by selecting the best varieties for a given farm plot) and more efficient product development efforts (e.g. by developing varieties for a specific farm or soil type). It can also provide timely and detailed advice to farmers on agronomic practices to improve their yields, using proprietary insights of the seed firm. Third, by incorporating digital farming into their portfolios, seed and agro-chemical firms can influence farmer choices by suggesting their own products to the detriment of competitors’ products or alternative methods (Philpott, 2016[59]).1313
Other developments within digital farming may evolve into substitutes for seed and agro-chemical inputs, for instance if precision farming can become sufficiently precise to reduce the benefits of broad-spectrum herbicides (such as glyphosate). As an illustration, Blue River Technology (acquired by John Deere in 2017 for more than USD 300 million) uses artificial intelligence and machine learning to allow the precise identification of individual weeds in a field. Combined with specialised machinery, this allows the precise application of herbicide to weeds only. These and similar technologies, if sufficiently advanced, could reduce the demand for herbicide tolerance traits.
At present, digital farming is still in its early stages and it is difficult to assess the potential complementarities and substitution effects at this point. However, interactions between digital farming and other agricultural input industries are likely to increase in importance in the future.
Intellectual Property Rights and access to genetic material
Complementarities have historically played an important role in stimulating non-horizontal mergers and acquisitions in the seed industry. At the same time, direct ownership by one firm is just one possible way of achieving these complementarities. Another possibility is by licensing the necessary intellectual property from other firms. For instance, a seed firm might license GM technology from a biotech firm, allowing it to include the biotech firm’s GM trait in its seed. Conversely, a biotech firm could license the use of proprietary germplasm from a seed firm. Similarly, exploiting the complementarity between herbicide and herbicide tolerance could occur through an ad-hoc collaboration between a chemicals company and a biotech firm. Mergers are not always necessary to exploit complementarity effects.
Major firms indeed cross-license genes, giving rise to GM seeds with genes from different seed firms. In 2007, for instance, Monsanto and Dow announced a collaboration to introduce SmartStax, an eight-gene stacked combination in maize. SmartStax combines several insect resistance and herbicide tolerance traits of both Monsanto and Dow. In terms of herbicide tolerance, SmartStax is compatible with both Monsanto’s RoundUp (glyphosate) herbicide as well as Bayer’s Liberty (glufosinate) herbicide. In addition to cross-licensing traits, the agreement between Monsanto and Dow also enables cross-licensing of germplasm of maize seed to achieve higher-yielding new hybrid combinations. By themselves, therefore, technological complementarities do not fully explain mergers and acquisitions, as firms seem capable of exploiting these complementarities through other organisational setups.
Other factors could explain why mergers and acquisitions have been one of the dominant ways by which companies exploited technological complementarities. First, given the large number of small seed firms and biotech firms initially active in the market, a strategy of licensing could have led to high transaction costs, especially since scientifically complex technical know-how is difficult to transfer between firms (Kalaitzandonakes and Bjornson, 1997[61]). Second, firms may also have used acquisitions as a way to ensure exclusive access to germplasm or traits, and to obstruct rivals’ access to these resources – or conversely, acquisitions may have been the only way to obtain access to rivals’ technological resources.
These potential explanations are intertwined within the complex role of Intellectual Property (IP) rights in the seeds and biotech industry. In parallel with the development of the seeds and biotech industry, IP rights on seeds and biotechnology have been strengthened over time (Box 3.1). The impact of stronger IP rights on mergers and acquisitions is unclear, however.
On the one hand, stricter IP protection may make it more difficult for firms to access each other’s technological resources, thereby making mergers and acquisitions more attractive. For instance, the breeder’s exemption in Plant Breeders’ Rights under the UPOV Convention allows firms to use each other’s germplasm for research and breeding purposes. In contrast, patents do not provide the same degree of freedom.1414 From this point of view, it cannot be argued that stronger IP protection makes it more difficult for firms to access each other’s genetic material, thus stimulating mergers and acquisitions as the best way to access protected IP. A similar argument holds for IP protection of other technologies, including GM traits and tools. Strong IP protection could lead to mutually blocking patent portfolios and potentially high transaction costs for cross-licensing technologies.
On the other hand, with stronger IP protection a firm may be more willing to license its IP as it faces a lower risk of IP theft or copycat behaviour by competitors. When intellectual property is not sufficiently protected, mergers and acquisitions may be the only way to safely combine intellectual assets as this removes the risk of theft of intellectual property or high litigation costs. Better protection of IP could therefore reduce the incentive for mergers and acquisitions as it allows firms to cross-license instead.
Plant breeding to develop a new variety is a costly and time-consuming process. Once a variety is introduced on the market, however, there is a risk that competitors and/or farmers will reproduce these seeds without having to incur the considerable R&D costs of the original breeder. Without a system of intellectual property rights, self-pollinating varieties such as wheat can easily be reproduced by competing seed firms (Fernandez-Cornejo, 2004[1]). Innovators would not be able to capture the full rewards for their efforts. This spillover problem may in turn reduce the incentives for private investments in R&D to create improved varieties. Two main policy solutions are investment in public R&D and/or the provision of intellectual property rights on new plant varieties to provide financial incentives for private R&D.
Intellectual property rights protection for new plant varieties has only emerged gradually. In most countries, intellectual property law initially did not allow for the protection of plant varieties. In the United States, for instance, the Patent Act of 1790 classified biological innovations such as new plant varieties as “products of nature,” which were excluded from protection (Fernandez-Cornejo, 2004[1]). Innovations by plant breeders could be freely reproduced by competitors and/or reproduced by farmers, reducing the incentive to invest in R&D.
Not all innovations in plant breeding suffered from this incentive problem to the same extent, however. Hybrid seeds in particular provide a biological protection against reproduction by both competitors and farmers. It is difficult to reproduce hybrid seeds, as this requires the parent lines from which hybrid seed is created. At the same time, offspring of hybrid seed is typically much less productive than the hybrid seed itself. As a result, farmers have less incentive to save seeds. The introduction of hybrid maize in the United States in the 1930s greatly increased the potential revenues to plant breeders of developing new maize varieties, even in the absence of intellectual property rights protection.
Over time, intellectual property rights were extended to cover plant varieties. The system of intellectual property rights for new plant varieties is mostly sui generis, that is, a system distinct from the patent system governing most other innovative industries. An international system was established by the International Union for the Protection of New Varieties of Plants (UPOV) through successive versions of its International Convention for the Protection of New Varieties. Countries which have implemented the UPOV Convention into national law protect new varieties using Plant Breeders’ Rights (PBR), also referred to as Plant Variety Protection (PVP) or Plant Variety Rights (PVR).1
A key characteristic of PBR is the “breeder’s exemption” which allows breeders to develop and commercialise new varieties from existing varieties without permission of the owner of the initial variety (with certain exceptions). The rationale behind this exemption is that “breeding is, by definition, the creation of improved varieties by recombining the characteristics of existing varieties” (ISF, 2012[62]). The goal of this exemption is to strike a balance between providing incentives for innovation and providing access to materials from which future plant varieties can be created.
In addition to PBR, plant-related inventions may be eligible for patent protection in some countries. In the United States, the Supreme Court ruled in Diamond v. Chakrabarty (1980) that genetically modified microorganisms can be patented. Since 1985, patents can also be used to protect new varieties in the United States, and plant breeders can use both a patent and a plant breeders’ right to protect the same variety (Pardey et al., 2013[63]). These patents do not have breeder’s exemptions. Not all jurisdictions are equally restrictive regarding patents on biological materials, however. In Switzerland, using patented biological material “for the purposes of the production or the discovery and development of a plant variety” does not constitute patent infringement.2 In the European Union, the “biotech directive” (Directive 98/44/EC) stipulates that a breeder in this situation could apply for a compulsory licence.3
The International Seed Federation, which represents plant breeders, is of the opinion that both PBR and patents are useful, but that PBR is the preferred form given the breeder’s exemption (ISF, 2012[62]). However, in the United States, seed companies often favour patents (Lence et al., 2016[64]).
While patents without breeder’s exemption may in theory slow the pace of innovation by making access to genetic material more difficult, they may provide stronger incentives for innovation. These two factors may need to be traded off against each other to decide on the optimal protection of intellectual property rights in plant breeding. Recent theoretical work suggests that patents may be more effective at stimulating innovation when research is expensive and long-lasting, while PBR may be more effective when improvements are expected to have a short life and research technology is easily transferable. Moreover, the same analysis indicates that a system of patents with effective and easy licensing might offer the best of both worlds in terms of offering incentives while maintaining access (Lence et al., 2016[64]).
An important policy issue regarding intellectual property rights for seed is the practice of farm-saved seed. Allowing farmers to save seed potentially reduces plant breeders’ revenues from developing a new variety, which would in turn reduce their incentives for innovation. The first UPOV Convention (1961) and the 1978 Act allowed governments to permit seed saving by farmers of varieties covered under PBR as long as it was not done for the production of seed for marketing.
These policy options were eventually clarified in the 1991 Act, which provides two exemptions. A first exemption allows for private and non-commercial uses by the farmer (e.g. subsistence farming). A second exemption allows countries adhering to the UPOV Convention to permit seed saving by farmers “within reasonable limits and subject to the safe-guarding of the legitimate interests of the breeder” (ISF, 2012[62]). In the United Kingdom, for instance, farm-saved seed is allowed but an industry-wide system exists for the collection of royalty payments on farm-saved seed. Royalty levels for farm-saved seed are lower than royalty rates on certified (“new”) seed. The system applies only to newer varieties (BSPB, 2014[65]).
_______________________________
1. In addition to plant breeders’ rights and patents, the United States also has a system of so-called plant patents, which cover only asexually reproduced plants (excluding potatoes and other tubers). Given their narrow scope, plant patents are not covered in this report; the term “patents” therefore always refers to the “regular” patent system (also known as utility patents in the United States).
2. Similar rules exist in France, Germany, and the Netherlands. Moreover, these rules are part of the Unified Patent Court agreement which will govern most European patents in the near future. At the time of writing, the agreement has been signed by all EU Member States except Spain and Poland. The OECD wishes to thank Marien Valstar for this information
3. For a global overview, see WIPO (2014[66]) and http://www.wipo.int/scp/en/exceptions.
Some critics allege that cross-licensing creates “non-transparent oligopolies” (Mammana, 2014[66]) or that these collaborations create “non-merger mergers” and raise questions about cartel behaviour (ETC Group (2008[67]), Howard (2009[49])). While it is possible that close collaboration facilitates explicit or implicit coordination, by itself cross-licensing is a pro-competitive practice as it allows competitors to access proprietary technology. In fact, a refusal to cross-license by a firm with a large portfolio of traits or germplasm would be more harmful for competition as this would force competitors to invest considerable amounts of money to develop their own traits or germplasm. To the extent that licensing and cross-licensing are possible with modest transaction costs and undertaken in a non-discriminatory way, the practice effectively reduces barriers to entry by eliminating the need to duplicate costly R&D programs. All else being equal, it reduces the incentive for firms to consolidate in order to obtain access to traits and germplasm, thus allowing more independent firms to remain active in the market.
Cross-licensing agreements, such as the SmartStax example mentioned above, are now common. In 2009, about 60% of the stacks on the US market involved traits from multiple firms (Moss, 2013[68]). The current IP regime in the United States therefore allows firms to collaborate through cross-licensing of intellectual property, rather than forcing them to engage in mergers or acquisitions to exploit these technological possibilities. At the same time, there are numerous examples of litigation between firms around supposed violations of intellectual property rights and licensing agreements. For instance, Oehmke and Naseem (2016[69]) argue that Dow’s 1998 acquisition of Mycogen may have been facilitated by the fact that Mycogen was embroiled in many costly legal battles over intellectual property rights, while Marco and Rausser (2008[70]) note that Monsanto’s acquisitions of both Calgene and DeKalb occurred in the midst of patent infringement suits. In recent years, leading vegetable seed firms have cooperated to create the International Licensing Platform – Vegetables to provide a clearinghouse for intellectual property rights with minimal transaction costs (Chapter 7).
Both theoretically and empirically, then, the impact of stronger IP protection on mergers and acquisitions in the seed and biotech industry is mixed. Similar conclusions were reached by Hall and Ziedonis (2001[71]), who studied the effects of stronger IP protection in the US semiconductor industry between 1979 and 1995. On the one hand, stronger patent rights appear to have facilitated the entry of specialised design firms focused on securing proprietary rights to technologies in niche product markets. This trend may also have contributed to non-horizontal disintegration in the industry. On the other hand, large firms engaged in a patent portfolio race, amassing vast patent portfolios to be used in litigation and negotiation with competitors. It seems plausible that firms with mutually blocking patent portfolios consider mergers and acquisitions as a way to overcome legal costs (Marco and Rausser, 2008[70]).
3.4. The evolution of markets over time: The case of US cotton
The historical evolution of the US upland cotton seed industry provides a good illustration of the structural changes discussed here. Between 1970 and 2017, the area planted with upland cotton in the United States varied between 4 and 6 million hectares. This market was one of the first to use genetically modified varieties. In contrast with most seed markets, public data is available on the market shares of different varieties (expressed as a share of acreage). These data stretch back several decades and allow for an assessment of long-term structural changes in the market. Moreover, Monsanto’s acquisition of a leading cotton seed firm (Delta & Pine Land) in 2007 provides an interesting case study on the impact competition authorities can have on the future evolution of seed and GM markets.
Structural changes in the US cotton seed industry
Figure 3.12 shows the estimated market shares of the main players in the US upland cotton seed industry between 1970 and 2017. In the 1970s, a high share of US upland cotton was provided by public plant breeders. Between 1970 and 1975, 16% on average of cotton planted were varieties from public institutes (with varieties from the University of New Mexico the most popular). Although at the time the market was quite concentrated, with four breeders controlling 74% of the market, this concentration has declined over time as Delta & Pine Land lost market shares; the four-firm concentration ratio reached a low of 52% in 1983. Starting in the 1980s, however, the US cotton seed sector witnessed several important structural changes.
First, the use of farm-saved seed declined strongly. Whereas purchased seed constituted only 50% of the total in 1982, this share increased to 75% by 1997 (Fernandez-Cornejo, 2004[72]). At the same time, the role of public plant breeding of US cotton became much less significant; by 1992, public institutes accounted for only 1% of the market. The Mississippi Agricultural Experiment Station had divested its popular DES-119 variety to Delta & Pine Land in 1991, while the market share of the Texas Agricultural Experiment Station fell due to the waning popularity of its flagship varieties, SP 21 and CAB-CS.
Second, Delta & Pine Land emerged as the clear market leader during this time thanks to successful new varieties such as DP 50, which by itself occupied 17% of the US cotton seed market by 1991. In 1994, Delta & Pine Land also acquired Paymaster and Lankart, further increasing its market share. By the year 2000, the estimated market share of Delta & Pine Land was almost 80%.
Third, in 1996 genetically modified cotton seed was introduced. Delta & Pine Land collaborated with Monsanto to introduce Bollgard (insect-resistant) traits into Delta & Pine Land’s cotton seeds. The resulting NuCOTN varieties were quickly adopted, reaching 17% of the total cotton seed market in 1997. As shown in Figure 3.13, GM cotton continued to spread, reaching 72% in 2000 and close to 100% in recent years.1515
Note: Showing main firms only. “Public” groups together public plant breeding research by the Agricultural Experiment Stations of Arkansas, Oklahoma, Texas, University of New Mexico, and Missisippi. Bayer includes Aventis CropScience (2000-2002). See main text for details on the ownership of Delta & Pine Land and Stoneville. Paymaster and Lankart were acquired by Delta & Pine Land in 1994.
Source: OECD analysis using United States Department of Agriculture – Agricultural Marketing Service, “Cotton Varieties Planted,” various years (1982-2017), and Fernandez-Cornejo (2004[72]) for 1970-1982.
Source: OECD analysis using United States Department of Agriculture – Agricultural Marketing Service, “Cotton Varieties Planted,” various years (1982-2017), and Fernandez-Cornejo (2004[72]) for 1970-1982.
Complementarities between germplasm, GM technology and agrochemicals led to mergers and acquisitions in the sector, leading to a complex set of changes in ownership. In 1997, Monsanto acquired Stoneville, and in 1998 Monsanto announced plans to merge with Delta & Pine Land. In preparation for this merger, Monsanto divested itself of Stoneville in 1999. However, the Department of Justice blocked the merger, leaving Monsanto without a strong position in cotton seed. Monsanto hence re-acquired Stoneville in 2005, but in 2006 undertook another attempt to acquire Delta & Pine Land, this time with more success. In 2007, Delta & Pine Land became a subsidiary of Monsanto, and Monsanto divested itself once again of Stoneville.
Bayer became active in the US cotton seed industry in 2002 after the merger with Aventis CropScience, and acquired AFD Seeds in 2005 and California Planting Cotton Seed Distributors (CPCSD) in 2007. That same year, Bayer acquired Stoneville from Monsanto.1616 Bayer’s cotton seed sales grew strongly during the early 2000s, to the detriment of Delta & Pine Land. Similarly, Americot has seen strong sales growth in recent years, a success which appears due in part to the NexGen breeding programmes it acquired from Monsanto in 2007 (Yancy, 2013[73]).
Historically, neither Dow nor DuPont have held strong positions in the cotton seed industry, but since the early 2000s Dow has played an increasingly active role through the PhytoGen Seed Company, a joint venture between Dow and the J.G. Boswell Company. PhytoGen sales have grown strongly since, reaching 18% of the market in 2012.
Innovation and product lifecycles
Innovation is a structural feature of the cotton seed market, and changes in firms’ market shares are closely linked with the success of new varieties. Moreover, the turnover of varieties appears to have accelerated in recent years. Figure 3.14 shows the product lifecycle of four cotton varieties introduced before the emergence of GM cotton. Of these, DP 50 and Acala 90 were introduced by Delta & Pine Land, while PM 145 and HS 26 were introduced by Paymaster. The success of the two Delta & Pine Land varieties (which had a joint market share of 29% in 1990) explains in large part the rise of this company in the late 1980s. The 1994 acquisition of Paymaster added the two other successful cotton varieties to Delta & Pine Land’s portfolio, although they had both peaked by that time.
Similarly, Figure 3.15 shows product lifecycles for successful cotton varieties launched after 1996. A striking difference is the shape of product lifecycles which seem to have been compressed. After the introduction of GM, new varieties appear to be adopted and abandoned faster. Whereas HS 26 took about seven years (between 1987 and 1994) to reach its peak, and another ten years to disappear from the market, Delta & Pine Land’s DP 555 BG/RR (known in the sector as “triple nickel”) took only three years to reach its peak after its introduction in 2002, and was already displaced by other varieties by 2011. Other varieties show a similar pattern of sharper increases and decreases in market share.
The main successful varieties before the mid-2000s were all products of Delta & Pine Land. By contrast, the late 2000s saw the emergence of varieties by Bayer (FibreMax, FM and Stoneville, ST) and PhytoGen (PHY). The growth in market share of these firms thus corresponds to the introduction of successful new varieties.
Note: Showing varieties introduced between 1982 and 1996 which obtained at least 10% market share.
Source: OECD analysis using United States Department of Agriculture – Agricultural Marketing Service, “Cotton Varieties Planted,” various years (1982-2017).
Note: Varieties introduced since 1996 which reached at least 10% market share before 2016. To improve visual presentation, one variety (PM 1218 BG/RR) is not shown, as its path nearly coincides with that of PM 2326 RR.
Source: OECD analysis using United States Department of Agriculture – Agricultural Marketing Service, “Cotton Varieties Planted,” various years (1982-2017).
The market for cotton GM traits
The concentration of GM traits is considerably greater than that of the cotton seed market itself (Figure 3.16). Since its introduction in the early 1990s, Monsanto’s insect-resistant BollGard traits and herbicide-tolerant Roundup Ready traits (including newer generations of this technology, such as Xtend) have been present in the majority of crop acres for most years.
In 2015, Bayer’s GlyTol and LibertyLink traits (both herbicide tolerance traits) gained popularity at the expense of Monsanto’s Roundup Ready traits, in line with Bayer’s growing share of the cotton seed market (Figure 3.12). As Bayer subsequently lost market share in cotton seed to Americot and Phytogen, its share of the GM traits market also fell as neither of these companies sell varieties that incorporate Bayer’s traits.
Dow’s WideStrike insect resistance traits have enjoyed a 12% to 18% share of US cotton acreage since 2010, as Phytogen (a Dow affiliate) incorporated WideStrike technology instead of Monsanto’s BollGard technology.
Syngenta has a presence in the market for GM cotton traits through its Vip3a insect resistance trait. This trait has been licensed to Bayer, Dow and Monsanto, and is included in these companies’ products TwinLinkPlus, WideStrike 3 and BollGard 3, respectively.1717 Combining the market shares of these products, Syngenta traits were present in 0.7% of cotton acres in 2016 and 3.7% in 2017.
Note: This figure shows estimated market shares of GM traits over time. Since traits are often stacked, shares sum up to more than one in most years. Figures aggregate different generations of the same technology (e.g. Roundup Ready, Roundup Ready Flex, Roundup Ready XtendFlex).
Source: OECD analysis using United States Department of Agriculture – Agricultural Marketing Service, “Cotton Varieties Planted,” various years (1995-2017).
Mergers and the role of competition authorities
The evolution of the US cotton seed market illustrates how decisions by competition authorities can shape the future evolution of a sector. In 2007, the Department of Justice allowed Monsanto’s acquisition of Delta & Pine Land, but imposed several conditions to address its concerns (Hogan Lovells, 2007[74]).
A first concern was that the merger would bring Stoneville (then owned by Monsanto) and Delta & Pine Land under the same owner, thus strongly reducing competition in cotton seed. A second concern was that post-merger, Monsanto could refuse access to its GM traits to competitors of Delta & Pine Land, thus reinforcing its strong position in cotton seed; and vice versa, Delta & Pine Land could refuse incorporating GM traits of Monsanto’s competitors, reinforcing Monsanto’s strong position in the GM traits market. The Department of Justice therefore required several remedies.
Monsanto was required to divest Stoneville as well as several Delta & Pine Land cotton varieties, additional Monsanto germplasm, and Monsanto molecular technology. Monsanto was also required to provide Stoneville with a license to Monsanto’s GM traits on the same terms as Delta & Pine Land had previously enjoyed. These assets were acquired by Bayer CropScience, with the exception of the NexGen brand which was acquired by Americot.
At the time of the merger, Syngenta and Delta & Pine Land were working to incorporate Syngenta’s VipCot insect resistance trait into 43 Delta & Pine Land cotton seed varieties. Monsanto was required to offer Syngenta the right to acquire these varieties to complete the work.
Monsanto was also required to revise its GM trait licenses to allow the stacking of Monsanto and non-Monsanto traits together.
In terms of market concentration in the seed market, these measures indeed seem to have prevented a single firm from obtaining a dominant position. Delta & Pine Land’s market share fell from 50% in 2006 to a low of 28% in 2014, although it recovered to 36% in 2017. Bayer’s acquisition of Stoneville immediately increased Bayer’s market share from around 30% to around 45%, and this share continued to grow to 50% in 2010. In the following years, Bayer would lose market share to other players, notably Americot. By 2017, Bayer’s market share declined to 14%, while Americot increased its market share from 12% in 2014 to 27% in 2017. As noted above, Americot’s recent success can be traced directly to the NexGen varieties it acquired from Monsanto during the 2007 divestitures. Phytogen, too, grew strongly after 2007. Whereas its market share was only 2-3% prior to the Monsanto-Delta merger, it increased to 18% in 2012. Hence, it appears that the measures put in place in 2007 indeed managed to maintain competition in the cotton seed market.
Note: Figure shows share of US cotton acreage by owner of GM traits. “Non-Monsanto only” includes Stoneville’s BXN Buctril 4EC herbicide-tolerant varieties (present in the market between 1997 and 2001); Bayer’s GlyTol, LibertyLink, GT, and GLT offerings (Bayer-only), Bayer’s GLTP offering (Bayer and Syngenta), and Dow’s WideStrike (Dow-only) offering. “Monsanto and others” includes stacks of Roundup Ready technology with Dow’s Widestrike technology (W3FE, W3F, WR, WRF) and stacks of BollGard technology with Bayer’s LibertyLink (LLB2, GLB2).
Source: OECD analysis using United States Department of Agriculture – Agricultural Marketing Service, “Cotton Varieties Planted,” various years (1995-2017).
The impact of the events of 2007 are also seen in the market for GM traits (Figure 3.17). Prior to 2007, nearly 100% of the available GM trait combinations were Monsanto only, with a combination of Monsanto’s BollGard insect resistance trait and Roundup Ready herbicide tolerance trait the most widely sold. After 2007, Monsanto was required to allow other firms to combine Monsanto and non-Monsanto GM traits. These combinations became more frequently available after 2007 and were planted on 35% of total cotton acres in 2014 and 2015. Likewise, varieties offering only non-Monsanto GM traits were practically non-existent in 2007. While these have remained a smaller part of the market, their share has increased over time, reaching 17% in 2016. In recent years, Monsanto-only GM traits have again increased their share, driven largely by the growth in Americot seed sales (which only incorporates Monsanto GM technology). While the structure of the market for GM traits hence clearly changed after 2007, Monsanto has maintained its leading position and other firms have found it difficult to displace Monsanto-only offerings.
3.5. Implications for the current merger wave
The literature reviewed here strongly suggests that innovation has been the main driver behind both horizontal and non-horizontal mergers and acquisitions over the past three decades. Firms sought to combine complementary intellectual assets (seed, traits, tools, and chemicals) through non-horizontal combinations. At the same time, R&D spending became a strategic tool for firms. As firms increased their R&D spending, the high fixed costs in turn stimulated horizontal consolidation.
These structural drivers are also at work in the current mergers. Statements by industry executives emphasize the importance of combining complementary product portfolios and the possibility of new complementarities arising from digital and precision agriculture (Bonny, 2017[12]).1818
The welfare effects of mergers depend strongly on whether a merger is non-horizontal or horizontal. Non-horizontal mergers are more likely to lead to efficiency gains (such as more efficient R&D through the combination of complementary assets), while for horizontal mergers there is a greater risk of negative effects on consumers through higher prices, lower innovation or fewer choices for consumers. At the same time, non-horizontal mergers are not necessarily harmless, as they may block competitors’ access to important resources – this was a particular concern in Monsanto’s 2007 acquisition of Delta & Pine Land and similar arguments apply to the recent mergers. The current mergers (in particular DowDuPont and Bayer-Monsanto) combine both horizontal and non-horizontal elements, and a key question for competition authorities is therefore to what extent efficiencies offset the risk of negative effects – and which measures need to be taken to reduce negative effects that might emerge. These arguments are reviewed in more detail in the next chapter.
Notes
← 1. In this report, the combined firm will be referred to as DowDuPont throughout as at the time of writing Corteva was not yet an independent business.
← 2. For example, ChemChina acquired the Italian tire maker Pirelli in 2015.
← 3. Some observers have speculated that the acquisition of Syngenta could help reduce Chinese consumers’ distrust of GM seeds, which have been virtually banned in China. This interpretation of the Syngenta acquisition is consistent with previous statements by President Xi Jinping, who declared in a 2013 speech that in order to secure China’s food security the country would need to “occupy the commanding heights of transgenic technology” instead of allowing foreign firms to dominate the market (The Economist, 2016[206]).
← 4. See also BASF (2017[209]), Marc Noel and Serafino (2017[208]) and Bray (2017[207]).
← 5. This geographic split likely reflects an important economic difference between field crop and vegetable seeds. Given their high value per kilogram, it is feasible to ship vegetable seeds from a single location. Field crop seeds, by contrast, are more bulky (i.e. have a lower value by weight), making it less efficient to source from a single location. This in turn means it is more difficult for a seed firm to sell field crop seeds in regions where it does not have an active physical presence (Marien Valstar, personal communication).
← 6. While not active in seeds, another important player is FMC, a leading supplier of crop protection chemicals. Following its acquisition of DuPont’s global pesticide business, FMC increases its sales of crop protection chemicals from USD 2.3 billion in 2016 to an estimated USD 4 billion in 2018. This makes FMC the fifth largest agricultural chemicals firm, after Bayer-Monsanto, ChemChina-Syngenta, DowDuPont and BASF.
← 7. One industry participant noted that this consolidation wave can be traced in part to the oil crises of the 1970s. Following high oil prices, players in the oil industry expected that energy production as well as molecules for plastic could in the future be derived from plants. This “green oil” trend led several petrochemical firms (e.g. Shell, Lubrizol and Elf) to invest in plant breeding and genetics firms.
← 8. Fixed costs are sunk when investments cannot (easily) be recovered once they have been made. Classic examples are investments in advertising or R&D. An important strand of literature on industrial organisation, following Sutton (1991[215]), explains empirical regularities in market structure with reference to such sunk costs. There is debate as to what extent theoretical arguments in this literature would also carry over to “non-sunk” fixed costs. In practice, most industries have a mix of sunk and non-sunk fixed costs (Sutton, 2007[54]).
← 9. Applications in the United States typically also include field trials, while most applications in the European Union were for “import and processing” purposes, for which data from field trials outside of the European Union can be used as input in the regulatory process.
← 10. For a detailed treatment of regulatory aspects of GM, including global governance and international trade aspects, see Smyth et al. (2017[251]).
← 11. The “life science” trend was not the only instance of a suspected complementarity which did not materialise. As noted earlier, some consolidation in the industry in the 1980s was inspired by the search for “green oil” (biological substitutes to fossil fuels).
← 12. In May 2018, digital agriculture was the topic of the OECD Global Forum on Agriculture. Ongoing OECD work is exploring the potential impacts of the digital revolution in agriculture, including implications for policy (Jouanjean and Deboe, 2018[240]).
← 13. For some firms, these digital offerings build on and extend agronomic services they were already offering to farmers (and which compete with public sector extension services). Advice based on digital farming may also crowd out independent crop consultants
← 14. Much of the discussion in the literature regarding the differences between plant breeders’ rights and patents has tended to focus on the US case, where plant varieties can be patented and where patent law does not provide a breeder’s exemption. Plant varieties cannot be patented as a whole in most jurisdictions, although biotechnology patents exist in many places. In the United States, use of such patented biotechnology in plant breeding does not benefit from a breeder’s exemption. However, this is not true for all jurisdictions (Box 3.1).
← 15. In collaboration with the United States Department of Agriculture, Delta & Pine Land itself had developed the controversial GM technology for genetic sterilisation of seeds known as “terminator genes.” Given strong public opposition, this technique has never been commercially implemented.
← 16. As part of the Bayer-Monsanto merger, Bayer is divesting its FiberMax and Stoneville cotton seed business to BASF
← 17. BollGard 3 varieties were not yet available in the 2017 season.
← 18. In addition to these underlying drivers, some other factors have been mentioned as explaining the timing of the current consolidation (Bonny, 2017[12]). First, after years of high crop prices creating favourable conditions for agricultural input industries, the decrease in agricultural prices in 2015-2016 negatively affected the seed and (especially) agrochemical industries. At the same time, the pesticide industry was experiencing tightening regulation in several countries. These factors may have encouraged industry participants to consider combinations with other firms. This tendency was further stimulated by low interest rates, which make it easier to finance large acquisitions.