Capabilities, Innovation and Industry Dynamics: Technological discontinuities and incumbents

Transcription

1 Capabilities, Innovation and Industry Dynamics: Technological discontinuities and incumbents Fredrik Tell KITE Research Group Department of Management and Engineering Linköping University

2 KITE Research Group For more information, please visit our website:

3 Three research themes: 1. Knowledge Integration and Project Organization 2. Knowledge Integration and Outsourcing 3. Innovation and the Integration of External Knowledge

4 Innovation, industrial dynamics and technological capabilities: Late shakeouts? Dynamics of complex capital goods industries Relationship between firm capabilities, innovation and performance Impact of firm capabilities on responses to (endogeneous) technical change Specific dynamics of mature phases in oligopolistic competition

5 What is an industry?

6 Product/Industry life cycles

7 Industry life cycles: Early shakeout patterns Innovation, entry and exit Competing explanations Exogenuous technology shocks (Jovanovic & MacDonald, 1994) Dominant designs (Abernathy & Utterback, 1975; Tushman & Anderson, 1986) R&D capabilities (Klepper, 1996; 1997; Klepper & Simons, 2002) => Favors early entrants (old and large firms) (Klepper, 1997; Klepper & Simons, 2005)

8 Exogenuous technological innovation Industries are created by initial inventions and shakeouts are triggered by subsequent refinement inventions (Jovanovic & MacDonald, 1994; Olleros, 1986) Basic invention => new product => entry => competitive equilibrium => entry ceases (shakeout) Refinement invention => new entry => incumbent firms at advantage in refining => expand output => fall in prices => non-innovators exit (shakeout) Favors early entrants (incumbents) No prediction on process innovations

9 Dominant designs A dominant design is a collection of enduring product standards to which the bulk of industry output eventually conforms (e.g. automobiles) (Abernathy & Utterback, 1978) Product architecture (Henderson & Clark, 1990) Initial uncertainty concerning designs many designs introduced => experimentation (product innovation) Network externalities and increasing returns to adoption induce convergence to a dominant design (Utterback & Suàrez, 1993, cf. David, 1985) Adopters (may) survive, non-adopters exit (shakeout) Processs innovations on dominant design lead to further concentration (shakeout) through returns to process innovations Favors early entrants (economies of scale)

10 Capabilities and increasing returns to R&D Scale advantages to R&D (Schumpeter, 1950) Firms conduct both product and process R&D Increasing returns to process R&D (Klepper, 1996) Product R&D returns independent upon pre-innovation level of output Process R&D returns favors firms with high output (proportional reduction of cost) Early entrants (and large entrants with related capabilities) are at favor and will survive continuous (no new equilibrium) shakeout driven by returns to R&D

11 Examples (Klepper & Simons, 2005)

12 Some observed empirical regularities Early concentration of entrants Prolonged shakeouts Early entrants came to dominate industries (Ford/GM; Goodrich/Goodyear/Firestone; RCA/Zenith/GE; Lilly/Wyeth/Squibb/Bristol/Pfizer) (Klepper and Simons, 1997; Klepper & Simons, 2005) Industry leaders dominated product and process innovation Pre-entry capabilities matter (Klepper & Simons, 2000; Helfat & Lieberman, 2002)

13 From the making of a oligopoly to the dynamics of oligopolies: Late shakeouts?

14 Technological capabilities and industrial dynamics in mature industries Technological capabilities and late shakeouts in the advanced gas turbine industry (Bergek, Tell, Berggren and Watson, 2008; Bergek, Berggren & Tell, 2009) Mature, but not a declining industry (cf Harrigan, 1980; Lieberman, 1985; Audretsch, 1995) rather growing

15 The example of Combined Combustion Gas Turbines (CCGT) Bergek, A., F. Tell, C. Berggren and J. Watson, (2008), Technological capabilities and late shakeouts: Industrial dynamics in the advanced gas turbine industry, , Industrial and Corporate Change, 17(2): Steam Generator Steam Steam turbine Generator ~ Fuel gas in Exhaust gases Intake Air Compressor Fuel Power Combustor turbine ~ Electricity Generator Advanced Turbine System

16 Combined Cycle Gas Turbines (CCGT)

17 Global trends in power generation Capacity (MW) CCGT Orders Total Orders

18 CCGT market growth Market development MW (yearly) GE 7F MW (cumulative) Market orders (yearly) Cumulative orders

19 Market share development GE 28 % 26% 22% 54% GEC-Alsthom /Alstom a 9% 14% 6% 15% ABB 18% 12% 17% Siemens 19% 24% 21% 22% Westinghouse 5% 7% 13% Mitsubishi b 13% 8% 12% 8% Other 8% 9% 9% 1% a GE licensee in the first three phases. In the fourth phase, Alstom acquired ABB s Power Generation Business. [i] b Westinghouse licensee in the first phases. [i] In 1989, the energy and transport businesses of Alsthom merged with GEC, forming GEC-Alsthom.

20 Two extremes: ABB and General Electric Million USD Revenues 25% Operating Profit Margins % % 10% % 0 0% ABB GE ABB GE Bergek, A., C. Berggren and F. Tell (2009), Do technology strategies matter? A comparison of two electrical engineering corporations, , Technology Analysis and Strategic Management, Vol. 21(4):

21 Research questions What were the characteristics of technological capabilities of the four major firms competing in CCGT? How did technological capabilities affect rates of innovation and, eventually, chances for survival in this segment of the electrical engineering industry?

22 How to explain the CCGT case? Industry life cycles? No exogeneous technology shock (Jovanovic and MacDonald, 1994) No product/process innovation pattern, (Abernathy and Utterback, 1978), continuous product development All firms were old and large (Klepper, 1996) Not a declining industry

23 Industry dynamics and product complexity Complex Products and Systems (CoPS) industries may remain in fluid phase, due to the architectural character of the product (Davies, 1997; Bonaccorsi & Giuri, 2000) Relatively stable firm structure, few exits and entries High entry barriers such as installed base, network externalities, and technological interdependencies Process innovations not as important in CoPS Specific technological capabilities (including intregration of new knowledge) pertaining to systems integrating (CoPS) firms (partly in line with Klepper)

24 CCGT as CoPS: Industry and firm characteristics Products Markets Manufacturing CoPS Mass production CoPS Mass production CoPS Mass production Many components Few components Oligopoly Competition High unit cost Low unit cost Systemic relationships Many alternative architectures Software/ control systems Analyzable relationships Few alternative architectures No component coordination Monopsony/ politicized purchasing Government regulation User-producer interaction Sophisticated buyer/operators Multitude of individual buyers Free markets Arms-length Relationship Non-professional buyers Customization Intensive technology Project-based organization Systems integration/ Breadth and depth Standardized Long-linked technology Functional organization Design-modularity/ Specialization (Magnusson, T., F. Tell & J. Watson (2005), From CoPS to Mass production? Capabilities and innovation in power generation equipment manufacturing, Industrial and Corporate Change, 14(1): 1-26

25 Technological capabilities: A simple conceptualization Technology Strategies Technology Activities Technological Capabilities Performance

26 Technological capabilities Technology strategies Technology leadership Cost focus Broad scope Technology sourcing Technology activities Product launching Patenting Problem-solving

27 Methodology Multiple measures and sources of data Annual reports Product launches and Relative market shares SPRU CCGT database on Power Plant orders Patents USPTO database (Linköping): Industry experts Thomson Derwent databases: Keyword search + manual code search Interviews and publicly available material (e.g., on sourcing and problem-solving)

28 Strategy measurements

29 TECHNOLOGY LEADERSHIP GE SIEMENS ABB WESTINGHOUSE 1987 X - Not available 1988 X X X Not available 1989 X x X X 1990 X x X - X X - X X 1993 X X X X X X X 1995 X X X X X X X - X X X X 1999 X X - Not available Not available 2001 X X - X = segment level statements; x = corporate level statements

30 Broad technology scope GE SIEMENS ABB WESTINGHOUSE (4) X (8) Not available (4) X (8) X (7) Not available (4) - (6) X (7) - (6) (4) - (7) X (7) - (6) (3) X (4) X (7) X (6) (3) - (6) X (8) - (5) (3) X (8) - (8) X (4) (4) - (6) X (8) - (4) (4) - (6) X (8) - (3) (5) - (5) X (8) - (4) 1997 X (4) - (5) X (7) - (4) 1998 X (4) X (7) - (9) 1999 X (2) - (6) (5) X (5) Not available Not available (3) X (5) (4) - (3) Note: All statements refer to the power generation segment. Numbers refer to the number of technology categories mentioned of 13 in total (see Appendix C).

31 COST FOCUS GE SIEMENS ABB WESTINGHOUSE 1987 X x Not available X Not available X X X x x X x - X 1995 X - X X x X Not available Not available

32 Technology sourcing GE: In-house + GE Aircraft Division The rest: In-house + collaboration

33 Product launch and sales impact Phase I Phase II Phase III Phase IV Siemens V94.3 W.house/MHI 701F ABB 13E2 ABB GT24 GE 7G, 9G, 9H (announced) W.house/MHI 501G Siemens V84.3A Orders (MW) GE Frame 7F W.house 501F GE GE licencees Siemens ABB Westinghouse MHI Other

34 Generation F and responses

35 Next generation

36 Total patenting Total number of patents, all searches combined (per application date) ABB GE Siemens Westinghouse

37 Technological capabilities: selected patents GE Siemens ABB Westinghouse Combined cycle ab 78 (5.2) 35 (2.3) 43 (2.9) 15 (1.0) Gas turbine engine (incl. measuring and testing) bc 865 (3.9) 685 (3.0) 220 (1.0) 227 (1.0) Gas turbines b 1031 (5.1) 293 (1.4) 204 (1.0) 217 (1.1) a Thomson keyword search, granted patents applied for b USPTO patent class search,granted patents applied for c Thomson manual code search, granted patents applied for

38 Distribution of patents 100% 90% 80% 70% 60% D % % 30% 20% 10% 0% ABB GE Entropy measure (E) = n i 1 P i ln(1/ P i ) 0 EABB 2.1 and 0 EGE 1.7 (cf. Zander, 1999, 2002)

39 Problems and the ability to solve them All manufacturers experienced serious problems in their installed plants, but they reacted quite differently: Public, Shipped home turbines, quick Public, Long problem-solving process Secrecy, continuing sales, slow, failure, Alstom Not much known, Mitsubishi

40 Technological capabilities strategies and activities

41 Some conclusions and further questions The importance of having a large and relevant capability base, built up by R&D activities, as a foundation for product development in complex technology fields. The study emphasizes the importance of integrating knowledge from several different technology fields in order to develop new architectural solutions on a sub-system level. How do firms access and integrate diverse technologies? A focused technology strategy on the segment level seems to be positively related to performance. Companies that focused on a limited number of technologies on the segment level were more successful than companies having a broad technology scope. Where to find the corporate coherence of large multi-technological corporations? The study shows that the development and launching of new products may not be as important as implicitly assumed in much of the capabilities literature, but rather solving after-launch problems proved more decisive for competitive outcomes. How do firms build problem-solving capabilities for emerging afterrelease problems?

42 Kowledge sourcing and integration Bergek, Tell & Palmberg (2010) study 41 alliances in the advanced gas turbine industry and relate modes of knowledge sourcing (organizational interdepence) with product architecture. Found that collaborative/open sourcing primarily takes place at sub-system level (and materials) (cf. Takeichi, 2001; Novak & Eppinger, 2001)

43 Conglomerates and coherence? Appliances Consumer electronics Lighting Media & Entertainment Aviation Power Generation Transportation Health care Materials Services Automation & Control Information & Communications Medical Power Generation Power Transmission Rail Transportation Services Automation Power Transmission Services

44 Explicit studies of coherence - findings Teece et al (1994), Journal of Economic Behavior & Organization Based on a statistical sample from 1987 on 18,620 diversified U.S. corporations larger than 20 employees, found that coherence defined as relatedness of neighboring activities remained constant as firms in the sample grew more diversified. Piscitello (2000), Structural Change and Economic Dynamics Studied patent data and product/business entry of 248 Fortune 500 firms between She suggested that the diversification strategies of firms were characterized by product-based coherence in the late 1970s to mid-1980s, and that strategies changed towards technology-based coherence from the late 1980s to the mid-1990s. Piscitello (2004), Industrial and Corporate Change Studied patent data (56 technological fields), product/business entry (42 sectors) and financial performance (after tax profits) of 248 Fortune 500 firms between She studied the corporate coherence as interconnectedness between technological and business diversification. She found some evidence that financial performance was positively influenced by companies abilities to increase corporate coherence between technologies and products. Breschi, Lissoni & Malerba (2003), Research Policy; in: Cantwell et al (2004) Investigated knowledge relatedness and coherence using a patent database covering all EPO patents for all firms from France, Germany, Italy, Japan, UK and the US. They found that relatedness in technologies is a major driver for firms technological diversification. They also found that large innovators were more coherent than smaller ones.

45 Some coherence studies cont d Nesta & Saviotti (2005), Journal of Industrial Economics; (2006) ICC Studied scope and coherence of U.S. pharmaceutical firms between 1990 and 1998, performing a citation analysis of 1,440 patents with 5,493 citations. They observe that knowledge base coherence contribute to innovative performance (and market value), and point to the importance of internal complementarities. Leten, Belderbos & Van Looy (2007), Journal of Product Innovation Management Using patent citation analysis of firm level patent data of 184 European, U.S. and Japanese firms active in five industries, found an inverted U- form relationship between technological diversification and technological performance. Their analysis also pointed to the moderating effect by technological coherence for the impact on technological diversification on technological performance, in the sense that net benefits of technological diversification are higher in technologically coherent portfolios. Bergek, Tell, Berggren, Watson (2008), ICC; Bergek Berggren & Tell (2009) Technology Analysis and Strategic Management Compared leading firms in the power generation equipment industry over a period of 15 years, using USPTO patent data, annual reports and financial performance. They found that GE, the most diversified firm on the corporate level was much more coherent (both in strategy and in technology) on the business level than ABB that was the least diversified firm on the corporate level in the sample. Moreover, GE had the best business performance in the industry.