Sunteți pe pagina 1din 7

Hormones, B. W. O'Malley and A. R. Means, 62. T. C. Spelsberg, W. Mitchell, F. C. Chytil, C. D. Liarakos, B. W. O'Malley, ibid., p.

Eds. (Plenum, New York, 1973), p. 1. B. W. O'Malley, Biochim. Biophys. Acta, in 2803.
42. J. Anderson, J. H. Clark, E. J. Peck, press. 85. J. Gorski, J. Biol. Chem. 239, 889 (1964).
Biochem. J. 126, 561 (1972). 63. F. C. Chytil and T. C. Spelsberg, Nat. New 86. T. H. Hamilton, C. C. Widnell, J. R.
43. J. H. Clark, 3. Anderson, E. J. Peck, Science Biol. 233, 215 (1971). Tata, Biochim. Biophys. Acta 108, 168 (1965).
176, 528 (1972). 64. T. Wagner and T. C. Spelsberg, Biochemistry 87. G. G. Maul and T. H. Hamilton, Proc. Nail.
44. , in Receptors for Reproductive Hor- 10, 2599 (1971). Acad. Sci. U.S.A. 57, 1371 (1967); R. G.
mones, B. W. O'Malley and A. R. Means, 65. G. C. Mueller, A. M. Herranen, K. J. Jervell, Roeder and W. J. Rutter, Biochemistry 9,
Eds. (Plenum, New York, 1973), p. 15. Recent Prog. Horm. Res. 14, 95 (1958). 2543 (1970).
45. S. Liao, J. L. Tymoczko, T. Liang, K. M. 66. J. Gorski, W. D. Noteboom, J. A. Nico- 88. J. Barry and J. Gorski, Biochemistry 10,
Anderson, S. Fang, Adv. Biosci. 7, 213 lette, J. Cell. Comp. Physiol. 66, 91 (1965). 2384 (1971).
(1971). 67. T. H. Hamilton, Proc. Nati. Acad. Sci. U.S.A. 89. W. L. McGuire and B. W. O'Malley, Bio-
46. T. C. Spelsberg, A. W. Steggles, B. W. 49, 373 (1963); ibid. 51, 83 (1964). chim. Biophys. Acta 157, 187 (1968).
O'Malley, J. Biol. Chem. 246, 4188 (1971). 68. A. R. Means and T. H. Hamilton, ibid. 56, 90. J. A. Nicolette and M. Babler, Arch. Bio-
47. B. W. O'Malley, T. C. Spelsberg, W. T. 1594 (1966). chem. Biophys. 149, 183 (1972).
Schrader, F. Chytil, A. W. Steggles, Nature 69. T. H. Hamilton, C.-S. Teng, A. R. Means, 91. M. Raynaud-Jammet and E. E. Baulieu, C.
(Lond.) 235, 141 (1972). ibid. 59, 1265 (1968). R. Hebd. Seances Acad. Sci. Ser. D Sci. Nat. P
48. G. Shyamala-Harris, Nat. New Biol. 231, 70. T. H. Hamilton, C. C. Widnell, J. R. Tata, 268, 3211 (1969).
246 (1971). J. Biol. Chem. 243, 408 (1968). 92. M. Arnaud, Y. Beziat, J. C. Guilleux, A.
49. D. 0. Toft, J. Steroid Biochem. 3, 515 (1972). 71. A. R. Means and T. H. Hamilton, Proc. Natl. Hough, D. Hough, M. Mousseron-Canet,
50. R. J. King and J. Gordon, Nature (Lond.) Acad. Sci. U.S.A. 56, 686 (1966). Biochim. Biophys. Acta 232, 117 (1971).
240, 185 (1972). 72. J. T. Knowler and R. M. S. Smellie, Biochem. 93. S. Mohla, E. R. DeSombre, E. V. Jensen,
51. D. 0. Toft, in Receptors for Reproductive J. 125, 605 (1971). Biochem. Biophys. Res. Commun. 46, 661
Hormones, B. W. O'Malley and A. R. Means, 73. D. N. Luck and T. H. Hamilton, Proc. Nati. (1972).
Eds. (Plenum, New York, 1973), p. 85. Acad. Sci. U.S.A. 69, 157 (1972). 94. M. Arnaud, Y. Beziat, J. L. Borgna, J. C.
52. K. R. Yamamoto and B. M. Alberts, Proc. 74. S. R. Glasser, F. C. Chytil, T. C. Spelsberg, Guilleux, M. Mousseron-Canet, Biochim. Bio-
Natl. Acad. Sci. U.S.A. 69, 2105 (1972). Biochem. J. 130, 947 (1972). phys. Acta 254, 241 (1971).
53. W. T. Schrader, D. 0. Toft, B. W. O'Malley, 75. A. B. DeAngelo and J. Gorski, Proc. Nati. 95. W. E. Hahn, R. H. Church, A. Gorbman,
J. Biol. Chem. 247, 2401 (1972). Acad. Sci. U.S.A. 66, 693 (1970). L. Wilmot, Gen. Comp. Endocrinol. 10,
54. W. T. Schrader and B. W. O'Malley, J. Biol. 76. T. H. Hamilton, C.-S. Teng, A. R. Means, 438 (1968).
Chem. 247, 51 (1972). D. N. Luck, in The Sex Steroids, K. W. Mc- 96. R. E. Rhoads, G. S. McKnight, R. T.
55. A. W. Steggles, T. C. Spelsberg, S. R. Glasser, Kearns, Ed. (Appleton-Century-Crofts, New Schimke, J. Biol. Chem. 246, 7407 (1971).
B. W. O'Malley, Proc. Nail. Acad. Sci. U.S.A. York, 1971), p. 197. 97. J. P. Comstock, G. C. Rosenfeld, B. W.
68, 1479 (1971). 77. C.-S. Teng and T. H. Hamilton, Proc. Natl. O'Malley, A. R. Means, Proc. Nail. Acad.
56. S. Liao, T. Liang, T. C. Shao, J. L. Ty- Acad. Sci. U.S.A. 60, 1140 (1968). Sci. U.S.A. 69, 2377 (1972).
moczko, in Receptors for Reproductive Hor- 78. R. H. Church and B. J. McCarthy, Biochim. 98. M. Mellie and J. 0. Bishop, J. Mol. Biol.
mones, B. W. O'Malley and A. R. Means, Biophys. Acta 199, 103 (1970). 40, 117 (1969).
Eds. (Plenum, New York, 1973), p. 232; 79. K. L. Barker and J. C. Warren, Proc. Natl. 99. S. E. Harris, A. R. Means, W. M. Mitchell,
W. I. P. Mainwaring and B. M. Peterken, Acad. Sci. U.S.A. 56, 1298 (1966). B. W. O'Malley, Proc. Natl. Acad. Sci.
Biochem. J. 125, 285 (1971). 80. G. Kapadia, A. R. Means, B. W. O'Malley, U.S.A., in press.
57. T. C. Spelsberg, A. W. Steggles, B. W. Cytobios 3, 33 (1971). 100. B. W. O'Malley, Biochemistry 6, 2546 (1967).
O'Malley, Biochim. Biophys. Acta 240, 888 81. A. R. Means, 1. B. Abrass, B. W. O'Malley, 101. A. R. Means and B. W. O'Malley, ibid. 10,
(1971). Biochemistry 10, 1561 (1971). 1570 (1971).
58. T. C. Spelsberg, A. W. Steggles, F. Chytil, 82. A. R. Means and B. W. O'Malley, Acta 102. G. Brawerman, J. Mendecki, S. Y. Lee,
B. W. O'Malley, J. Biol. Chem. 247, 1368 Endocrinol. Suppl. 153, 318 (1971); R. D. ibid. 11, 637 (1972).
(1972). Palmitter, A. K. Christensen, R. T. Schimke, 103. G. C. Rosenfeld, J. P. Comstock, A. R.
59. P. 0. Kohler, P. M. Grimley, B. W. O'Malley, J. Biol. CI em. 245, 833 (1970). Means, B. W. O'Malley, Biochem. Biophys.
J. Cell Biol. 40, 8 (1969). 83. R. J. Britten and D. E. Kohne, Science 161, Res. Commun. 47, 387 (1972).
60. S. H. Socher and B. W. O'Malley, Dev. Biol. 529 (1968). 104. B. W. O'Malley, G. C. Rosenfeld, J. P.
30 (2), 411 (1973). 84. C. D. Liarakos, J. M. Rosen, B. W. O'Malley, Comstock, A. R. Means, Nat. New Biol. 240.
61. C. A. Perotta, Am. J. Anat. 111. 195 (1962). Biochemistry 15, 2809 (1973); J. M. Rosen. 45 (1972).

sizes of the samples are summarized in


Table 1. A simple concept of the fac-
tors that limit and determine a firm's
effectiveness in innovation and of the
phases and relationships in the innova-
tive process is presented below; this con-
Innovation in Industry cept allows one to compare the find-
ings from these diverse sources.
and the Diffusion of Technology In this article, I present what we
know-or think we know-about the
process of innovation by firms. How
James M. Utterback do characteristics of the environment
affect firms' innovation? What factors
and information affect the creation and
acceptance of ideas for new products?
What factors are related to effective
The impact of technological change innovate and to spend greater amounts development efforts? What do we
on economic growth (1), industrial on research and development, and know about the acceptance of innova-
productivity, and international compe- whether and how to reduce the bar- tions in the market and about the
tition and trade has been widely recog- riers to innovation faced by firms (5). creation of new firms based on tech- -
nized. There is a rapidly developing A wealth of hypotheses and case stud- nology? Finally, past work in this field
interest in such issues as the environ- ies of the process through which tech-
ment for advances in science (2) and nology is created, developed, and used The author is associate professor, Graduate
technology (3) and the contribution by firms is available and should pro- School of Business, Indiana University, Blooming-
and relationship of basic science to vide a useful perspective in dealing ton 47401. This article is adapted from a paper
commissioned by the National Scien9e Foundation
technology (4). Recent debate has fo- with these questions. The sources of Task Force on Alternate Federal Policies Affect-
more than 2000 case studies, the indus- ing the Use of Science and Technology, directed
cused on the questions of whether and by J. Herbert Hollomon (NSF grant GQ-5),
how to provide incentives to firms to tries or innovations studied, and the June to December 1971.

620 SCIENCE. VOL. 183


appears to have been of a distinctly There is a striking similarity between from sources outside the firm. Mueller
descriptive and noncumulative nature. the findings of studies conducted in the (18) found that 14 of DuPont's 25
What are the crucial issues for research United States and those conducted in major product and process innovations
and how might these be approached the United Kingdom (10-12). originated wholly outside that com-
to yield more scientific, rigorous, and Innovation also appears to be stimu- pany. Of 59 pieces of information in-
cumulative results? lated by expanding markets and by corporated in the ideas for 32 new
Any firm's potential for technical rising costs of inputs, with innovations scientific and measuring instruments,
innovation can be considered as a aimed at reducing the use of more ex- 39 came from outside the firm that
function of its environment-including pensive inputs (13, 14). Firms tend developed the idea (19). Langrish
economic, social, and political factors, to innovate primarily in areas where (20) found that 102 of 158 key ideas
the state of development of technol- there is a fairly clear, short-term po- involved in generating 51 innovations
ogy, and information about technol- tential for profit (15). Many innova- came from outside sources.
ogy. Barriers to flows of people and tions of great commercial significance If one examines the innovations
information between the firm and its are of the relatively low-cost, incre- named by firms as commercially suc-
environment will limit its knowledge mental type, the result largely of con- cessful, one finds that a significant
of social and market needs, new and tinuous development efforts (7, 16, number (23 to 33 percent) have been
existing technology, and government 17). wholly adopted from other firms.
programs, incentives, and regulations, In most industries, no single firm These are more often process innova-
thus limiting the potential for innova- commands a majority of the resources tions than product innovations, and
tion as seen by the firm. Character- available for research, nor can any one tend to be modifications rather than
istics of the firm itself, including its firm respond to more than a portion completely new items. Interestingly,
resources, personnel, and patterns of of the needs or problems requiring the cost of the adopted innovations
communication and decision-making, original solutions. It is not surprising, was about the same as that of the
will determine the degree to which it therefore, to find that most of the original innovations studied (7). This
meets its perceived potential for inno- ideas successfully developed and im- similarity in cost is probably due to the
vation (6). plemented by any firm came from out- fact that the cost of originating and
Innovation, as distinct from an in- side that firm. Of the 157 cases stud- developing a successful innovation is
vention or technical prototype, refers ied by Myers and Marquis (7) for a minor part (probably 15 to 30 per-
to technology actually being used or which these data are available, 98 of cent) of the total cost of bringing it
applied for the first time. The process the ideas were evoked by information into use (21).
of innovation is viewed, for simplicity
in making comparisons, as occurring
in three phases: generation of an idea, Table 1. Some retrospective studies of technological innovation.
problem-solving or development, and
implernentation and diffusion. Genera- Author Industries studied Sample
tion of an idea involves synthesis of size
diverse (usually existing, as opposed Arthur D. Little, Inc. (27) Textiles 12
to original) information, including in- Machine tools 6
Construction 8
formation about a market or other Semiconductors 12
need and possible technology to meet Carter and Williams (11) 116 British firns 204
the nieed. Problem-solving includes Enos (14) Petroleum refining (processes only) 11
setting specific technical goals and de- Other industries 35
signing alternative solutions to meet Goldhar (37) Winners of Industrial Research Award 108
them. Implementation consists of the Hamberg (22) Major innovations 1946-1955 27
manufacturing-engineering, tooling, and Jewkes et al. (23) Major innovations 1900-1945 61
plant and market start-up required to Langrish (12) British innovations given
Queen's Award in 1966 and 1967 51
bring an original solution or invention Mansfield (30) Iron and steel 49
to its first use or market introduction. Petroleum refining 66
Diffusion takes place in the environ- Bituminous coal 28
ment and begins after the innovation Railroads 10
Miller (24) Steel products and processes 12
is introduced (7-9). Mueller (18) DuPont's major innovations 1920-1949 25
Myers and Marquis (7) 9 railroads 79
14 railroad equipment suppliers 125
53 housing suppliers 196
-Environmental Factors 12 computer manufacturers 90
23 computer equipment suppliers 77
NSF-lIT (34) Magnetic ferrites
Market factors appear to be the pri- Video type recorder
-mary influence on innovation. From Oral contraceptive pill 5
60 to 80 percent of important innova- Electron microscope
Matrix isolation
tions in a large number of fields have Peck (15) Aluminum 194
been in response to market demands Robertson et al. (10) Chemicals 34
and needs. The remainder have origi- Scientific instruments 24
nated in response to new scientific or Sherwin and Isenson (17) Weapons systems 20
technological advances and opportuni- Tannenbaum et al. (33) Major materials developments 10
ties. These data are shown in Table 2. Utterback (31) Instruments 32
15 FEBRUARY 1974 621
Larger firms do not seem to develop Table 2. A comparison of studies of the pro- (31) and by Goldhar (37). A;pproxi-
a greater proportion of innovations, portions of innovations stimulated by market mately 40 percent of the respondents
needs and technological opportunities.
relative to their market share, than in each of these samples hold the
smaller firms (18, 22-24). No con- Propor-
tion
Propor- Ph.D. degree. Personal contacts, edu-
tion cation, and experience constitute by
sistent relationship between size of frm
firm and number of innovations ap- from far the largest proportion of informa-
Autor
Author o ' nical Sample
tech- sze tion sources used in originating ideas
peared in the Myers and Marquis (7) or pro- oppor-
study except, perhaps, a stronger duction tunities for innovations. The data from these
market orientation on the part of small- needs studies suggest that education is the
er firms. Mansfield (25) suggests that primary avenue through which basic
size of firm has little effect on inno- Baker scientific findings are translated into
et al. (38) 77 23 303* engineering practice.
vation, at least when a fim is above
Carter and
some threshold size. Structural factors Williams (11) 73 27 137
affect this relationship from industry Goldhar (37) 69 31 108
to industry. Shimshoni (26) notes that Sherwin and Sources of Ideas
those smaller firms which are success- Isenson (17) 61 34 710t
ful innovators in the instrument in- Langrish (20) 66 34 84 In most cases, ideas for innovations
dustry rely largely on government con- Myers and originate with communication about a
Marquis (7) 78 22 439 need, followed by search for technical
tracts and orders early in their life Tannenbaum
cycle, later diversifying into commer- et al. (33) 90 10 10 possibilities to meet the need (19, 38).
cial areas. In mature industries, such Utterback (31) 75 25 32 Informal and oral sources provide the
as textiles, machine tools, and con- * Ideas for new products and processes. t Re-
majority of key communications about
struction, innovation is more likely to search events used in 20 developments. both needs and technical possibilities
come from smaller, new firms than (7, 20, 38). Communication about a
from older, large firms, as well as need seems often to be initiated by
from firms in other industries (27). mentioned above (32). Tannenbaum someone other than the person who
This generalization also appears to be (33), Sherwin and Isenson (17), and generated the idea for an innovation,
true of the petroleum refining industry the National Science Foundation (34) while communication about a technical
(28). have undertaken studies of the con- means is initiated most often by the
There is a substantial lag, 8 to 15 tributions of basic and applied research innovator himself (19).
years, between the time technical in- to innovation. These studies have used Consultants, consulting activity, and
formation is generated and the time widely differing criteria for selecting information resulting from diversity in
it is used in an innovation. The lag the innovations studied, for defining work assignments appear to play ma-
appears to vary with industry, product, and analyzing time lags, and for choos- jor roles in the generation of ideas for
market, and resources used. Enos (14) ing the time periods to be included, successful innovations. For example,
concludes that mechanical innovations and are not, therefore, directly com- outside experts played a crucial part
have the shortest interval, with chem- parable. The TRACES study (34) in the generation of ideas for 16 of the
ical and pharmaceutical innovation suffers from having forced an overly 32 new instruments I studied (9).
next, and electronic innovations taking rational pattern on the data obtained Peters (39) has explored the relation-
the most time. In addition, he states -assuming that information which ships among consulting, diversity in
that "the interval appears shorter appeared relevant in retrospect was work assignments, and generation of
when the inventor himself attempts to actually a factor in the progress of a ideas in interviews with faculty in four
innovate than when he is content given innovation (35). Both the departments at the Massachusetts In-
merely to reveal a general concept" TRACES and Hindsight (17) analyses stitute of Technology (MIT). He
(14, p. 309). The lag appears to be note that applied research and devel- found that 96 percent of those re-
shorter for innovations directed at opment concerns often stimulated fur- porting ideas engaged in consulting,
consumer, as opposed to industrial, ther basic study. This point is sup- as opposed to 55 percent of those not
markets and for innovations developed ported also by data reported in the reporting ideas. Of those reporting
by government, as opposed to those Materials Advisory Board study (33). ideas, 70 percent said also that their
from industry (29). Because of the These data strongly suggest that the work was mixed between research and
small samples studied, the above find- crucial role of basic research in indus- development, as opposed to 28 percent
ings are of questionable validity. Fi- trial innovation lies in continual rein- of those not reporting ideas. Gordon
nally, the time required to develop and forcement and understanding of the and Morse (40) also note that con-
bring an idea to first use appears to implications of applied work. sultation outside the work setting tends
be relatively constant, with the median Where data on the individuals in- to enhance generation of ideas. These
between 1 and 7 years for various volved in generating successful inno- findings might be- explained by the
samples (29-31). vations are reported, the conclusion is central requirement for synthesis of
Basic research does not seem to be that they are a well-educated group; information in forming ideas. Both
significant as a direct source of inno- however, all levels of education are consulting and diversity in work as-
vations. It plays a critical role in the represented. The median education of signments would tend to assist in
production of knowledge and enters the founders of new firms, Roberts bringing together information ahbout
the process of innovation indirectly, (36) reports, is the master's degree. needs and about technical possibilities.
by means of education. This role is This is also true for the originators of As noted above, in a majority of
partly responsible for the time lag innovations reported by Utterback cases an idea results from recognition
622 SCIENCE, VOL. 153
of a need, followed by a search for the result of communication initiated technical gatekeepers account for a
relevant technology. In some cases, by the innovator. Myers and Marquis majority of the ideas for solutions
however, recognition of a technical (7) note that 17 percent of all infor- noted as outstanding by Allen's re-
opportunity stimulates the search for mation that evoked the basic ideas for spondents. It would be difficult to de-
an application of the new technology the innovations they studied was ob- sign a more effective system of infor-
or information. Older technical possi- tained at the initiative of others. This mation dissemination than the one he
bilities seldom attract attention spon- finding was true for only 3 percent of describes.
tarteously; in contrast, a new discovery the cases in which the information was Allen (45) contends that the com-
or technical possibility might well at- used in problem-solving. In contrast, plex nature of technical information
tract attention and stimulate a search 12 percent of the information that and of individual user needs provides
for applications. Thus, one might ex- evoked ideas was the result of search a strong incentive for the development
pect innovations stimulated by a need by the innovators, as opposed to 25 of such a mediating activity in the
to be based on older technology than percent during problem-solving. (The flow of technical com'munication. The
innovations stimulated by a technical magnitudes of these differences are not lack of success in attempts to design
opportunity. This surmise has been comparable because of the different computer-based systems for retrieval of
supported by several studies (7, 19, categories used in the two studies.) technical information (46) is not diffi-
37, 41). How can the latest technical I have found (19) that oral sources cult to understand in light of these
information be used in meeting needs? were important both during idea-gen- findings. In general, one can conclude
Periodic retraining of technical per- eration (45 percent of all informa- that increased communication outside
sonnel and concentration on informal tion) and during problem-solving (32 of the immediate work group and or-
communication, personal mobility, and percent of all information). While in- ganization will, other factors being
diversity in work and consulting op- formal sources still played a major equal, be related to better performance
portunities could serve to reduce the role during problem-solving, the pro- in problem-solving and that increased
discrepancy between available technol- portion of inputs from primary sources communication could be achieved
ogy and technology in use. (analysis and experiment) was found more effectively with policies designed
The above findings may well explain to double from 22 percent during idea- to encourage the development of in-
the fact that government-held patents generation to 52 percent during prob- formal channels of communication (4)
and technical reports are seldom used lem-solving. These data all point to a or, at least, designed not to impede
in a commercially or socially impor- more active and structured search for this process.
tant application other than the spe- information during problem-solving
cific one from which the patent or in- (32).
formation arose. Since application of Most of the information used in Internal Characteristics of the Firm
technology is usually stimulated by a problem-solving comes from within the
need or -market, one would not expect firm (7, 19). However, this informa- Barriers to communication and ac-
the availability, per se, of patents or tion is usually brought into the firm tion within the firm and its resources,
technical information to result in ap- by a few individuals, termed "techni- organization, and other internal char-
plication. The findings above do imply cal gatekeepers" (43), who have more acteristics were noted as limiting the
that more commercially oriented pat- extensive contact than the others do firm's ability to originate, develop, and
ents would find greater application, as with colleagues outside the firm or implement innovations in response to
would patents or information gener- with technical literature, or both. These communication with its environment.
ated in agencies or firms with diverse persons are frequently chosen as in- In general, firms face not one strategy
missions or markets and thus with a ternal consultants or technical infor- but a rich variety of possible strategies
greater chance to couple the new mation sources by others within the for dealing with a given set of oppor-
technology with needs. Recent data firm. The findings suggest that, while tunities and problems. Each possible
support these expectations (42). highly developed, internal technical strategy is associated with a cost, and
resources and communication channels each is more or less appropriate in
are vital to success in problem-solving, different environments.
Sources of Solutions to Problems information flow from the environ- As the uncertainty faced by the firm
'ment is also critical to effective tech- increases, its need for specialization to
Communications that aid in generat- nical solutions. This appears to be deal with varying facets of its environ-
ing ideas are often initiated by some- more the case for rapidly changing ment-such as market, production,
one other than the person who has the technologies than for stable fields. and technological factors-also in-
idea. In contrast, information that is Allen (44) has shown not only that creases. Uncertainty may arise from a
important in developing ideas usually technical gatekeepers have much great- number of sources, such as the clarity
comes from communications initiated er contact than others with professional of information available, the extent
by the person involved in solving de- literature and oral sources of informa- of knowledge about causal relation-
velopment problems related to the idea. tion outside the firm, but also that ships among environmental factors,
I have found (19) that roughly half they are typically in close communi- and the length of time required to
of all information that stimulated inno- cation with others in the organization
vations came from communication ini- judge the impact of any environmental
who share these characteristics. Infor- change or management action. Regard-
tiated by someone other than the in- mation tends to be communicated less of the levels of specialization, the
novator himself. However, during quickly within the gatekeeper network need for integration among functions
problem-solving, fully 86 percent of and from the gatekeepers to others in appears to remain relatively constant;
the important information used was the firm. It is not surprising that the however, the difficulty and cost of
15 FEBRUARY 1974
623
achieving integration among functions straints. In more stable technical fields, in either a product or a process lead-
will increase as specialization increases the need for specialization would be ing to a reduction in the average total
(47). This increase is even more pro- less, and therefore the disadvantage in cost of production per unit. Relative
nounced when a firm faces rapid, terms of impairment of technical per- advantage may also be the result of
short-term fluctuations in its environ- formance of a total project organiza- increased demand for the finished pro-
ment, perhaps as a result of govern- tion would be reduced (45). duct because of improved product
ment actions, and must resort to tem- Marquis (53) found that projects quality or variety, which leads to in-
porary expedients to achieve needed in- having some slack resources not only creased total revenue. Finally, relative
tegration (48). Inappropriate mana- achieved better cost and schedule per- advantage may result when an innova-'
gerial responses, such as demand for a formance, virtually by definition, but tion allows an increase in price and
highly structured organization in a also tended to have better technical thus a higher average revenue per unit.
rapidly changing environment, or use outcomes than those without slack. Another factor involved in the adop-"
of too few, poorly placed, excessively This finding underscores the generally tion decision is the degree of associated
costly or excessively permanent inte- adverse relationship between time and risk because of the absolute cost of an
grating devices, will usually be asso- cost, on the one hand, and technical innovation, its cost relative to the
ciated with poor performance by the quality, on the other. Achieving a firm's resources, and the ability and
firm (47, 49). given technical advance in a reduced willingness of the firm to absorb the
Organizational and spatial bonds period of time generally results in costs of a wrong decision (56). Rising
might be expected to affect communi- much higher costs. Some devices, such aspirations, based on increased sales,
cation and integration among func- as PERT (Program Evaluation and profitability, and market share, as well
tions and between phases in the inno- Review Technique), designed to alle- as expanding markets, may also en-
vation process (50). For example, viate this problem have aided in im- courage adoption. However, Gold's
transfers of technical personnel among proving communication among supe- analysis (57) of the diffusion of 14
divisions, other factors being equal, riors and subordinates, but they have major innovations in the steel industry
would be expected to result in a tem- not necessarily resulted in fewer cost fails to support this proposition. Al-
porary increase in communication be- overruns or fewer delays. The cost of though he found no cases of rapid dif-
tween the divisions. This increase innovation will clearly be lower, and fusion without rapid growth in output,
would be particularly likely if the per- the chance of effective technical per- both medium and slow adoption rates
son transferred was a part of, or was formance greater, if needless environ- were associated with all categories of
linked to, the gatekeeper network in mental uncertainties can be avoided growth rates, including zero and nega-
one or both organizations (51). Re- or reduced, because the resources re- tive rates. Neither does relative, ad-
tention of a liaison person or group as quired for integration will be corre- vantage offer an explanation of why
a project moves from idea-generation spondingly less. and under what conditions firms will
through problem-solving to implemen- seek higher profits by adopting innova-
tation has been suggested as an effec- tions rather than by choosing other al-
tive strategy for integration (52). Diffusion of Innovations ternatives.
Architecture exerts a significant impact The rate of diffusion of an innova-
on communication, with frequency of Diffusion of innovations in the tion can be measured with respect to
communication falling off very rapidly market is considered to be a two-step the percentage of firms that have
with increasing physical barriers and or multistep flow similar to the gate- adopted the innovation, or with respect
distance among people (44). keeper phenomenon described above to the percentage of total output ac-
These factors have been examined (54). In consumer markets, diffusion counted for by the innovation. Diffu-
by Marquis (53) in an extensive study begins slowly, with a few influential sion rates appear to depend on infor-
of the relationship between organiza- individuals' use of the new product or mal and personal communication for
tional structure and project success. process; their experiments initiate much the same reasons that communi-
Organization of technical personnel by wider communication and use of the cation about technology during prob-
function was found to be related to innovation (55). These propositions lem-solving does. The information in-
more effective technical performance, have been explored fully in cases volved is complex, buyers have varving
while organization of administrative where individuals or families are the needs, and information and needs tend
personnel by project was related to purchasers; there is a much smaller to change continually, requiring a flex-
more effective cost and schedule per- body of research on adoption decisions ible communication linkage.
formance. Marquis concludes that a by firms. Early awareness of an innovation
hybrid, or matrix, organization, in The probability that a given firm appears to depend on external sources,
which there is a small project team but will adopt a product or process is such as advertising and vendors (58).
where more than half of the technical thought to be an increasing function Evaluation and adoption, however,
personnel remain in their functional of the proportion of firms in the seem to depend to a greater extent on
departments, is the best option. This industry already using it and of the communication with technical person-
type of organization is more likely to profitability of doing so, but a de- nel inside the adopting firm. At each
meet needs for both specialization and creasing function of the size of the stage in the adoption process, the
integration in an uncertain environment investment required (25). The rela- amount, quality, and value of informa-
than either a total project or a total tive advantage afforded by an innova- tion available appear to have an im-
functional organization; thus, this type tion seems to be the primary determi- pact (59). Factors that may tend to
of organization is more likely to achieve nant of whether or not it is adopted retard diffusion include the degree to
technical excellence and, at the same in an industrial market. Relative ad- which an innovation is incompatible
vantage may be the result of a change with existing processes and requires
time, to meet cost and schedule con-
SCIENCE, VOL. 183
624
major process changes, the degree to ventures formed by technical entrepre- further research, but do not appear to
which increased technical skills are re- neurs leaving established firms can be offer a means for deeper understanding
quired to use the innovation, and the seen from data gathered on spin-offs of the innovation process. The retro-
probability that major improvements from the MIT Lincoln Laboratory. spective nature of nearly all of the
will rapidly alter the innovation, mak- The 50 new ventures formed prior to sources discussed probaably means that
ing delay in adoption advantageous the study had resulted in a total em- the process has been viewed as much
(56, 57). ployment greater than that of the Lin- more rational and well-ordered than it
No relationship was found between coln Laboratory itself, a constant 1800 is in fact. This failing is partially over-
the size of a firm and its relative abil- people. Similarly, 36 ventures that were come in firsthand accounts such as
ity to innovate. Similarly, there is no spin-offs from a large Boston electron- those of Suites and Bueche (63) and
evident relationship between firm size ics firm over a 5-year period had total Frey and Goldman (64). Each of these
and speed of adoption of innovations. sales exceeding those of the parent firm accounts involves a successful innova-
Larger firms appear to lead in some at the end of the period. The better- tian according to technical or commer-
industries, while smaller and medium- performing half of 84 firms studied by cial criteria, or both. However, many
sized firms lead in others. Nor does Roberts (36) exhibited a high degree of the characteristics of innovations
leadership in adoption appear to be of technology transfer from the parent that have failed commercially (10) ap-
concentrated in particular firms in the organization. pear to be similar to those of success-
few industries for which data are avail- There is some evidence that spin-offs ful cases. The few longitudinal studies,
able. Webster (56) contends that larger are a function of the environment both and studies comparing more and less
firms, more able to afford the new in- inside the firm itself and ,in its geo- successful cases, do support the main
vestment required for adoption and graphic area. For example, spin-offs conclusions drawn above (10, 32, 38).
more able to tolerate the risk of adop- tend to occur in areas of application More serious problems are raised by
tion, will adopt innovations earlier. that have large number of potential the distinctly nonrepresentative nature
However, smaller firms are more likely parent firms and organizations, as op- of the samples used. There are few
to value the technology involved in posed to attractive new areas. The ma- cases (17, 33, 65) in which the con-
adoption and to have less complex de- jority of entrepreneurs tend to have a tributions of more than one organiza-
cision-making processes, which may development orientation rather than a tion, or details of interactions over a
lead to earlier adoption. research orientation and generally significant period of time, are discussed.
One clear implication of the above tended to be frustrated by constraints There is a wide variation in the impor-
findings is that there is a significant or other factors in their previous jobs tance of the innovations included, rang-
time lag between the appearance of a (62). Some evidence indicates that the ing from those affecting the economy
new technology and any wide eco- recent flux in scientific and technical as a whole to cases involving produc-
nomic and social impact of its use. employment has resulted in a relatively tion in a single firm, albeit with signifi-
While firms may adopt technology for greater formation of new enterprises. cant comniercial results (66).
their own short-term advantage, moni- This finding may be tempered by a In addition to questions of compara-
toring the initial outcome of such higher failure rate for these firms. bility and sampling, a central problem
adoptions should provide a means for Many of the spin-off companies for further research on innovation will
determining long-run social and eco- studied began as government contrac- be to devise an operational model to
nomic consequences. Where adoption tors or as sellers of products for de- account for interfirm and interindustry
of an innovation is judged desirable, fense and space purposes. In time, differences. Polar definitions used in
this adoption may be encouraged by these companies usually increase their past studies, "high technology" and
incentives designed to increase the rela- sales to commercial markets (36). "mature industry," for example, are
tive advantage to be gained from adop- This mode of entry is clearly dominant insufficient.
tion or to reduce the risks associated for new firms in the scientific instru- One possibility is to use the strategy
with adoption. ment industry (26, 31). These new for growth or competition evident in a
firms also typically involve a high de- firm or an industry, such as sales maxi-
gree of advanced technology and tech- mization (automotive), cost minimiza-
New Firms Based on Technology nology that was transferred from other tion (transportation, communications),
organizations and developed to pene- performance maximization (aircraft,
In addition to output, skilled persons trate new markets and areas of applica- chemicals), or control of materials re-
and technical information flow from a tion. sources (m,ining, petroleum), as a
firm and become involved in the crea- basis for drawing distinctions (67). For
tion of additional innovations. This example, in an industry that seeks to
type of flow from technically based Conclusions maximize sales, one would expect in-
firms may have a greater effect on em- novations that would be highly visible
ployment and the economy than the The varied definitions used in the to consumers to be developed rapidly
activities of the parent firm itself. The sources that have been discussed make (68). In a cost-minimizing situation,
spin-off ventures formed are character- any aggregate analysis difficult. A sim- production, as opposed to product
ized by high growth and survival rates ple three-stage analysis of flows to, technology, would be a major source
and by a high degree of technology from, and within the firm was used of uncertainty, while the reverse might
transfer into new markets. Roberts to facilitate comparisons. Even so, each be the case in a performance-maximiz-
(60) and Cooper (61 ) have studied of the generalizations is drawn from ing situation. Greater uncertainty aris-
such transfers and spin-offs in the Bos- relatively small and unrepresentative ing from technical sources would imply
ton and Palo Alto areas, respectively. samples. Case studies may continue to greater sophistication in effective firms'
The economic and social impact of new be a source of ideas and hypotheses for product planning approaches, while a
15 FEBRUARY 1974 625
more stable technology would imply 4. W. J. Price and L. W. Bass, Science 164,
802 (1969).
Science Foundation, lllinois Institute of Tech-
nology Research Institute, Chicago, 1968).
greater sophistication in market re- S. G. Black, What's Happening to Small Busi- 35. B. Leach, Scientfic Techniques and Scinee
search and market-oriented strategies ness R & D (George Washington Univ., Policy (Univ. of Manchester Business School,
Washington, D.C. 1971); J. Hollomon, Manchester, England, 1971).
for innovation, and so forth. Much Technol. Rev. 74, 10 (September 1972). 36. E. B. Roberts, Res. Manage. 11, 249 (July
*more work is needed along these lines 6. J. M. Utterback, Acad. Manage. 14, 75 (March 1968).
1971); B. D. Estafen, ibid., p. 51. 37. J. D. Goldhar, thesis, George Washington
if outcomes of interventions in the in- 7. S. Myers and D. G. Marquis, Successful University (1970).
Commercial Innovations NSF 69-71 (Na- 38. N. R. Baker, J. Siegman, A. H. Rubenstein,
novative process are *to be predicted tional Science Foundation, Washington, D.C., IEEE Trans. Eng. Manage. EM-14, 156 (De-
with any accuracy. 1969).
8. W. H. Gruber and D. G. Marquis, Factors 39.
cember 1967).
D. H. Peters, thesis, Massachusetts Institute.
Some implications for providing in- in the Transfer of Technology (MIT Press, of Technology (1968).
centives and reducing barriers do seem Cambridge, Mass., 1969). 40. E. V. Morse and G. Gordon, Acad. Manage.
9. J. M. Utterback, in Technological Change Proc. 28th Annu. Meet. (December 1968), p.
clear from the work to date. Effective and Economic Growth, G. W. Wilson, Ed. 37.
directions for federal action lie in strat- (Indiana Univ., Division of Research, Bloom- 41. J. M. Utterback, "Factors affecting innovation '
ington, 1971), pp. 139-160. in industry and the diffusion of technology,"
egies such as creating new markets 10. A. B. Robertson, B. Achilladelis, P. Jervis,
Success and Failure In Industrial Innovation:
paper prepared for the National Science
Foundation (1971).
through purchases or procurement poli- Report on Project Sappho (Centre for the 42. Harbridge House, Inc., Government Patent
cies; aggregating or focusing markets Study of Industrial Innovation, London, 1972). Policy Study: Final Report (Government
11. C. F. Carter and B. R. Williams, Industry Printing Office, Washington, D.C., 1968).
through regulation and other means; and Technical Progress: Factors Governing 43. T. J. Allen, Technol. Rev. 70, 2 (October-
providing for market entry by contracts the Speed of Application of Science (Oxford November 1967); and S. I. Cohen.
Univ. Press, London, 1957). Adm. Sci. Q. 14, 12 (March 1969).
to smaller firms, venture capital, strong- 12. J. Langrish, M. Gibbons, G. Evans, F. R. 44. T. J. Allen, R & D Manage. 1, 14 (October
er patent protection, and so on; and Jerons, Wealth from Knowledge (Macmillan, 1970).
New York, 1972). 45. , in Annual Review of Information,
providing for mobility and informal 13. J. Schmookler, Invention and Economic Science, and Technology, C. A. Cuadra, Ed.
Growth (Harvard Univ. Press, Cambridge, (Encyclopedia Britannica, Chicago, 1969), pp.
contacts within the technical commu- Mass., 1966). 3-31.
nity. Technology "push" strategies 14. J. L. Enos, in The Rate and Direction of 46. S. Doctors, The Role of Federal Agencies in
Inventive Activity: Economic and Social Technology Transfer (MIT Press, Cambridge.
(such as tax 'incentives) to increase Factors, R. R. Nelson, Ed. (Princeton Univ. Mass., 1969).
most research spending, prizes for new Press, Princeton, N.J., 1962), pp. 299-322. 47. J. W. Lorsch and P. R. Lawrence, Orga-
15. M. Peck, in ibid., pp. 278-298. nization,and Environment: Managing Differ-
technology, and documentation and in- 16. S. Hollander, The Source of Increased Ef- entiation and Integration (Harvard Business
ficiency: A Study of DuPont Rayon Plants School, Division of Research, Boston, 1967).
formation retrieval systems would prob- (MIT Press, Cambridge, Mass., 1965). 48. J. R. Galbraith, in Studies in Organization
ably be less important in stimulating 17. C. W. Sherwin and R. S. Isenson, Science Design, J. W. Lorsch and P. R. Lawrence,
Eds. (Irwin, Homewood, Ill., 1970), pp. 113-
156, 1571 (1967).
innovation. 18. W. F. Mueller, in The Rate and Direction 139.
Definitive answers will require the of Inventive Activity: Economic and Social 49. T. Burns and G. M. Stalker, The Manage-
Factors, R. R. Nelson, Ed. (Princeton UniO. ment of Innovation (Tavistock, London, 1961).
most difficult kind of research-experi- Press, Princeton, N.J., 1962), pp. 323-360. 50. J. R. Morton, Organizing for Innovation: A
ments in the field. Since the interven- 19. J. M. Utterback, IEEE Trans. Eng. Manage. Systems Approach to Technical Management
EM-18, 124 (1971). (McGraw-Hill, New York, 1971).
tions required are difficult and expen- 20. J. Langrish, R & D Manage. 1, 133 (June 51. P. G. Gerstberger, thesis, Massachusetts In-
1971). stitute of Technology (1971); T. J. Allen,
sive in most cases, they will not be 21. U.S. Department of Commerce, Technological Technol. Rev. 73, 36 (March 1971).
under the researcher's control. Nor will Innovation: Its Environment and Management 52. A. H. Rubenstein and C. F. Douds, IEEE
(Government Printing Office, Washington, Trans. Eng. Manage. EM-16, 137 (November
the effect of policy changes be visible D.C., 1967); W. J. Abernathy and R. S. 1969).
over a short period. Thus it seems im- Rosenbloom, Manage. Sci. 15, 486 (June 1969). 53. D. G. Marquis, Innovation 1, 26 (July 1969).
22. D. Hamberg, J. Polit. Econ. 71, 95 (April 54. J. S. Coleman, E. Katz, H. Menzel, Medical
perative to take advantage of interven- 1963). Innovation: A Diffusion Study (Bobbs-Merrill,
tions that occur fortuitously to con- 23. J. Jewkes, D. Sawers, R. Stillerman, The Indianapolis, Ind., 1966).
Sources of Invention (Macmillan, London, 55. E. M. Rogers, Diffusion of Innovations (Free
struct "quasi-experiments" (69) with 1958); B. Johannisson and C. Lindstrom, Press, New York, 1962); and F. F.
Swed. 1. Econ. 73, 427 (October-November Shoemaker, Communication of Innovation
as great a degree of control over other 1971). (Free Press, New York, 1971).
factors as possible. For example, have 24. R. E. Miller, Innovation, Organization and
Environment (Univ. of Sherbrooke, Sher-
56. F. E. Webster, Jr., J. Mark. 33, 35 (July
1969).
recent changes in policy regarding fed- brooke, Quebec, 1971). 57. B. Gold, Pitts. Bus. Rev. 41, 6 (1971).
erally held patents increased the com- 25. E. Mansfield, Industrial Research and Tech- 58. F. E. Webster, Jr., J. Mark. Res. 7, 186 (May
nological Innovation: An Econometric Anal- 1970).
mercial use of these patents? Have ysis (Norton, New York, 1968). 59. , ibid. 5, 426 (November 1968).
changes in the capital gains laws re- 26. D. Shimshoni, thesis, Harvard University 60. E. B. Roberts and H. A. Wainer, in Pro-
(1966). ceedings of the 20th National Conference on
tarded the development and growth of 27. Arthur D. Little, Inc., Patterns and Problems the Administration of Research (Denver Re-
of Technical Innovation in American Industry search Institute, Denver, Colo., 1966), pp.
"spin-off" enterprises? Has the identi- (report C-65344 to the National Science Foun- 81-92.
fication of technology gaps (3) and dation, Arthur D. Little, Inc., Cambridge,
Mass., 1959).
61. A. C. Cooper, IEEE Trans. Eng. Manage.
EM-18, 2 (February 1971).
competitive opportunities stimulated in- 28. G. Foster, Process Innovation in Petroleum 62. , R & D Manage. 3, 59 (February
novation? The effects of such actions (Sloan School working paper 490-70, MIT, 1973).
Cambridge, Mass., 1970). 63. C. G. Suits and A. M. Bueche, in Applied
on technical innovation could be care- 29. F. Lynn, Report of the National Commission Science and Technological Progress, H. Brooks
fully observed with a modest but sus- on Technology, Automation, and Economic et al., Eds. (Government Printing Office,
Progress (NCTAEP, Washington, D.C., 1966). Washington, D.C., 1967), pp. 297-346.
tained research effort, which promises 30. E. Mansfield, The Economics of Techno- 64. D. N. Frey and J. E. Goldman, in ibid., pp.
to yield valuable information beyond logical Change (Norton, New York, 1968). 255-295.
31. J. M. Utterback, thesis, Massachusetts In- 65. C. Layton, Ten Innovations (Crane, Russak,
that available from largely historical stitute of Technology (1969). New York, 1972).
D. G. Marquis, Innovation 1, 28 (November
32. T. J. Allen, thesis, Massachusetts Institute of 66.
sources. Technology (1966). 1969).
33. M. Tannenbaum et al., Report of the Ad Hoc 67. W. H. C. Simmonds, Technol. Forecast.
References and Notes Committee on Principles of Rese,rch-Engi- Soc. Change 4, 375 (April 1973).
neering Interaction (report No. MAB 222-M, 68. L. J. White, The Automobile Industry Since
l. E. Mansfield, Science 17S, 477 (1972). National Academy of Sciences-National Re- 1945 (Harvard Univ. Press, Cambridge, Mass.,
2. K. W. Deutsch, J. Platt, D. Senghaas, search Council Material Advisory Board, 1971).
ibid. 171, 450 (1971). Washington, D.C., 1966). 69. D. T. Campbell and J. C. Stanley, in Hand-
3. Organisation for Economic Co-operation and 34. Illinois Institute of Technology Research In- book of Research on Teaching, N. L Gage,
Development, The Conditions for Success in stitute, Technology in Retrospect and Critical Ed. (Rand-McNally, New York, 1963), pp.
Technological Innovation (OECD, Paris, 1971). Events in Science (report to the National 171-246.

626 SCIENCE. VOL. 183

S-ar putea să vă placă și