Title: Same As it Ever Was: The Search for Hypercompetition
1BADM 551, MT1Law, Technology and Intellectual
Property
Paul M. Vaaler Associate Professor of
International Business Department of Business
Administration College of Business University of
Illinois at Urbana-Champaign
Session 12 Competition Policy and IPM
2Are Technology-Intensive Industries More
Dynamically Competitive? No and Yes
Paul M. Vaaler Associate Professor of
International Business College of
Business University of Illinois at
Urbana-Champaign, Illinios, USA pvaaler_at_uiuc.edu
Gerry McNamara Associate Professor of
Management Broad School of Management Michigan
State University, East Lansing, Michigan,
USA mcnamara.gerry_at_msu.edu
3Research Question Background
- Fundamental Strategy Question How Do We Explain
Differences in Firm Performance? - IO Economics Stable industry factors (Bain,
1956 Porter, 1979) - Corporate Strategy Stable corporate portfolio
benefits and managerial actions (Chandler, 1962
Montgomery, 1979) - Business Strategy Stable, business-level
idiosyncratic resources and positioning effects
(Porter, 1980 Wernerfelt, 1984 Barney, 1991) - Hypercompetitive Perspective Unstable markets,
entrepreneurial advantage, adroit management of
fluid, short-term factors in increasingly
volatile markets (DAveni, 1994 Thomas, 1996
Hamel, 2000)
4Increasingly Hypercompetitive Industries
- Driven by Exogenous and Endogenous Factors,
Industries Have Become more Hypercompetitive
(DAveni, 1994, 1995) - Within These Industries, There Will Be
- Increased boldness in action, including attacks
on industry leaders and the redefinition of
markets and products - More frequent and intense competitive
interactions - Increased emphasis on speed in strategic actions
- Focus on mobile and or substitutable resources
- Outcome Firms are Increasingly Less Able to
Establish Sustainable Advantages. Instead, they
Increasingly Act to Create Temporary Advantages
and Erode the Advantages of Rivals
5Prior Findings
- Scant Broad-Based Evidence of Increasing
Hypercompetition - Supporting case study, specific industry, short
time period research on dynamic industries
(DAveni, 1994, 1995 Robins Wiersema, 2000
Rindova Kotha, 2001) - Less supporting evidence of any broader
hypercompetitive shift. Some evidence
suggesting shift in pattern of rivalry and
profitability in select manufacturing industries
(Thomas, 1996 Wiggins and Ruefli, 2005) - Evidence of increasing dynamism not confirmed in
a broad set of industries (Castrogiovanni, 2003
McNamara, Vaaler, Devers, 2003) - But What About the Possibility that Increasing
Hypercompetition Is a More Limited Phenomenon?
6The Role of Technology and the Most Affected
Industries
- Researchers in Management, IO and Antitrust
Economics Have Identified Technology as a Key
Driver of Increased Dynamism in Markets - New Competitive Landscape Driven by a
Technological Revolution (Bettis Hitt, 1995) - Technological change leading to a
neo-Schumpeterian era where firms need to
innovate continually (Garud Karaswarmy, 1995) - Heightened turbulence in information,
communications, and other technology-related
industries (Chakravarthy, 1997) - New processes for high-velocity,
technology-driven industries (Brown Eisenhardt,
1998) - Industries experiencing rapid changes in
contextual factors such as technology (Bogner
Barr, 2000)
7The Role of Technology and the Most Affected
Industries
- IO and Antitrust Economics and Increasing
Dynamic Competition - Schumpeterian dynamic competition (Schumpeter,
1912 1942) - Innovation-based competition threatening to sweep
away leading firms, transforming industries and
advancing national economic development - Recent IO and antitrust views (Technology-based
competition generating red queen effects
(Khalil, 1997 Posner, 2000 Schmalansee, 2000
Ahlborn, Evans, Podilla, 2001 Evans
Schmalansee, 2001 Baumol, 1993 2002) - Modern dynamic competition Technology-based
competition for markets (rather than price and
output-based competition within markets) - Implications Knowledge-based competition
battles over tech standards episodic winner
take all or most battles for markets
temporarily extremely profitable winners
red-queen effects where firms have to innovate
faster just to keep up - Policy responses Less antitrust oversight
certify industries as high-tech and have no
antitrust oversight belief in the
self-equilibrating market
8The Central Research Question
- Do We See Evidence of the Consequences of
Increasing Hypercompetition in Technology-Intensiv
e Industries (Dynamic Competition)? - Evidence of increasing competitive dynamics over
time (longitudinal evidence in TI industries) - Evidence of higher competitive dynamics relative
to other industries (cross-sectional evidence
comparing TI to non-TI industries)
9Logical Consequences of Hypercompetition in TI
Industries
- Higher General Industry Dynamism
- Lower Persistence of Abnormal Performance
- Lower Persistence of Market Share Leadership
- Higher Mortality (Exit) Rates
10Increasing General Industry Dynamism
- Rapid Rise, Maturity, and Obsolescence of
Technologies Leads to Greater Industry Dynamism
(Chakravarthy, 2001) - Rapid Role Out, Imitation, and Replacement of New
Products and Services Results in Boom and Bust
Cycles in Industries (Schumpeter, 1912 Agarwal
Gort, 2001 McKnight, Vaaler, Katz, 2001) - H1a TI industries will exhibit greater dynamism
over time. - H1b TI industries will exhibit greater dynamism
than non-TI industries.
11Decreasing Persistence of Abnormal Performance
- Lower Barriers to Entry to and Imitability within
Industries (DAveni, 1994) - Deregulation
- Capital market changes
- Disintermediation due to technology advances
- Increased use of alliances
- Increase in Technological Change and Diffusion
(Garud Karaswarmy, 1995) - Shorter Product Life Cycles and Quicker Imitation
(Agarwal Gort, 2001) - Increasingly Frequent Winner Take Most Battles
(Evans Schmalansee, 2001) - H2a Abnormal business returns in TI industries
will erode more quickly over time. - H2b Abnormal business returns in TI industries
will erode more quickly than for firms in non-TI
industries.
12Decreasing Persistence of Market Share Leadership
- Lower Barriers to Entry to Industries (DAveni,
1994) - Shorter Product Life Cycles and Quicker Imitation
of First Movers (Agarwal Gort, 2001) - More Frequent and Bolder Attacks on Industry
Leaders Requiring Response (or Leading to
Dethronement) (DAveni, 1994 Ferrier et al.,
1999) - Increasingly Frequent Winner Take Most Battles
(Evans Schmalansee, 2001) - H3a The likelihood that market leading firms in
TI industries are supplanted from one year to the
next will increase over time. - H3b The likelihood that market leading firms in
TI industries are supplanted from one year to the
next is higher than for leading firms in non-TI
industries.
13Increasing Firm Mortality Rates
- Dramatic Environmental Changes Resulting in
Sudden Obsolescence of a Firm Resources and
Threatening Survival (Hannan Freeman, 1977,
1984 Tushman Anderson, 1986) - Technological Change Resulting in Industry
Over-Supply and Increased Firm Mortality
(Christensen, 1997) - Winner Take Most Battles Resulting in Large
Number of Distressed Firms (Posner, 2000, 2001) - H4a Mortality rates for firms in TI industries
will increase over time. - H4b Mortality rates for firms in TI industries
will be greater than for firms in non-TI
industries.
14Research Methods
- Sample Drawn from Compustat Industry Segment
Database 14,000 obs from 2800 BUs Operating in
31 Industries with High RD Intensity Covering
1978-1997 Period - Hypotheses Evaluated Two Ways 1) TI Industries
Alone (Longitudinal) and 2) TI Industries
Compared to Non-TI industries (Cross-Sectional)
- Analytical Models Hypotheses Testing
Predicted Trend - OLS Regression General Industry
Dynamism Positive - Autoregressive Model Persistence of Abnormal
Returns Negative - Logistic Regression Persistence of Above
Average Returns Negative - Logistic Regression Persistence of Market
Leadership Negative - Hazard Rate Model Mortality (Industry Exit)
Rate Positive
15Industry Dynamism Analysis
- Analytical Model Hypotheses Testing
Predicted Trend - OLS Regression General Industry
Dynamism Positive - Dependent Variable Variability in industry
sales over a five-year window (longitudinal) or
over the entire study period (cross-sectional)
(Dess Beard, 1984) - Control Variables Industry dummy variables
(longitudinal analysis) - Hypothesized Variables
- Time period dummy variables (longitudinal
analysis) - TI industry dummy variable (cross-sectional
analysis)
16Industry Dynamism Results
- Results No significant positive time trend in
TI industries, nor differences between TI and
non-TI industries. But adding two additional
controls (average industry sales, number of years
reported in Compustat database) yields evidence
of cross-sectional differences (TECHi 0.012 and
t-value is 77.14).
Compared to Non-TI Industries
TI Industries Over Time
Variables
Time Period 1 (1978-1982) -0.0281
(0.0604) Time Period 2 (1983-1987)
0.0028 (0.0485) Time Period 3 (1988-1992)
0.0383 (0.0485) TI Industry (TECHi)
-0.0152 (0.0095)
17Autoregressive Analysis
- Analytical Model Hypotheses Testing Predicted
Trend - Autoregressive Model Persistence of Abnormal
Returns Negative - Dependent Variable ROA in year t (ROAjt)
- Control Variables ROA in year t-1 (ROAjt-1),
industry structure (HHI), GDP growth, inflation,
time counter, TI industry dummy (cross-sectional) - Hypothesized Variables
- Interaction of year counter variable and lagged
ROA (longitudinal analysis) - Interaction of TI industry dummy variable and
lagged ROA (cross-sectional analysis)
18Autoregressive Results
Results No evidence that abnormal returns are
less persistent in TI industry businesses over
time. Compared to non-TI industry businesses, TI
businesses have lower returns, but increasingly
more (positive) stable year-to-year performance.
TI Industries Over Time
Compared to Non-TI Industries
Variables
Year Interaction (ROAjt-1YEAR)
-0.0005 (0.0014) TI Industry Dummy
(TECHi) -0.0140 (0.0012) TI Industry
Interaction (ROAjt-1TECHi) 0.0524 (0.0
033)
19Performance Persistence Analysis
- Analytical Model Hypotheses Testing Predicted
Trend - Logistic Regression Persistence of High
Returns Negative - Dependent Variable Variable indicating whether
or not a business sustained a certain performance
level from one year to the next (Ruefli
Wiggins, 2005) - ROA above industry average
- ROA 1 standard deviation above industry average
- ROA 2 standard deviations above industry average
- ROA 3 standard deviations above industry average
- Control Variable Industry structure (HHI), GDP
growth, inflation (as before) - Hypothesized Variables
- Year counter variable (longitudinal analysis)
- TI industry dummy variable (cross-sectional
analysis)
20Performance Persistence Results
- Results Mixed. Above average business returns
more persistent in TI Industries, but the
sustainability of very high business returns
lower in TI Industries and over time. - Above-Average
- Variables Performance 1 STD 2 STD
3 STD - TI Industries over time
- Year Counter (YEARt ) 0.0452 0.0145
-0.0744 - (0.0083) (0.0147) (0.0379)
- Compared to Non-TI Industries
- TI Industry (TECHi) 0.1082 -0.0514
-0.2684 -0.8635 - (0.0341) (0.0596) (0.1450)
(0.3123)
21Persistence of High Performance of TI Firms
22Market Leader Persistence Analysis
- Analytical Model Hypotheses Testing
Predicted Trend - Logistic Regression Persistence of Market
Leadership Negative - Dependent Variable Variable indicating whether
or not business with the largest industry sales
base in one year sustained it in the next year - Control Variable Industry concentration (HHI)
- Hypothesized Variables
- Year counter variable (longitudinal analysis)
- TI industry dummy variable (cross-sectional
analysis)
23Market Leader Persistence Results
- Results No significant negative time trend.
Also, no evidence that market share leadership is
more difficult to sustain in TI industries
compare to other industries.
Compared to Non-TI Industries
TI Industries Over Time
Variables
Year Counter (YEARt) -0.0083 (0.0204) TI
Industry (TECHi) -0.0782 (0.1079)
24Market Leader Persistence Over Time
25Mortality Analysis
- Analytical Model Hypotheses Testing Predicted
Trend - Hazard Rate Model Mortality (Industry Exit)
Rate Positive - Dependent Variable Variable indicating whether
or not business in existence in one year remained
in existence in the following year - Control Variables Industry density, Density2,
Value of MA activity in year t - Hypothesized Variables
- Year counter variable (longitudinal analysis)
- TI industry dummy variable (cross-sectional
analysis)
26Mortality Results
- Results No significant positive time trend.
Also, no evidence that mortality is greater in TI
industries compare to other industries.
Compared to Non-TI Industries
TI Industries Over Time
Variables
Year Counter (YEARt) -0.0035 (0.0023) TI
Industry (TECHi) -0.0075 (0.0095)
27Mortality Rate Over Time
28Research Conclusions and Implications
- Is Dynamic Competition Greater in TI Industries?
No (and maybe) Yes. - No Support for the Assertion of Broadly
Increasing Hypercompetition within TI Industries - Some Support for the Assertion that TI Industries
Are Significantly More Hypercompetitive than
Non-TI Industries - Some (a Little) Support for Assertion of
Increasing Hypercompetition within TI Industries,
but Only in a Narrow Strata of Very hHigh (But
Not High or Highest) Performing TI Industries - A fringe theory to explain increasingly
challenging conditions for winning firms in TI
industries? - We should know better
29Limitations and Future Research
- If Only We Had Done Another Test
- If Only We Had Done the Same Tests in Another
(More Recent) Time Period Or With The Right
Sample - If Only We Had Thought About Dynamic
Capabilities - For the Future Closer Look At Causal Chains
linking industry conditions, firm actions, and
performance consequences for top performing firms - Moving in and out of periods of heightened
competitive dynamics - Causes of differential impact at varying
performance levels - For the Future The Role of Cognition in the
Dynamics of Industries - Managerial perceptions
- Perceptions of other stakeholders (e.g.,
investors and analysts)
30Are Technology-Intensive Industries More
Dynamically Competitive? No and Yes
Thank You!
Paul M. Vaaler Associate Professor of
International Business College of
Business University of Illinois at
Urbana-Champaign, Illinios, USA pvaaler_at_uiuc.edu
Gerry McNamara Associate Professor of
Management Broad School of Management Michigan
State University, East Lansing, Michigan,
USA mcnamara.gerry_at_msu.edu