Transcript Slides

Henderson, R. & Cockburn, I. (1994). Measuring
Competence? Exploring Firm Effects in Pharmaceutical
Research. Strategic Management Journal, 15: 63-84.
Seung Hoon Lee
Overview
Empirical work on Resource Based View so far…
Firm effects account for diversification strategy (Hitt &
Ireland, 1985; Montgomery & Wernerfelt, 1988) and
performance (Cool & Schendel, 1988; Rumelt, 1991).
But…
Few studies considered particular competences
(had to rely on aggregate level measures of competence)
And more importantly, a puzzle remained from their own
recent research (using the same data set).
- A more direct motivation of the study
Henderson, R. & Cockburn, I. (1994). Scale, Scope and Spillovers:
The Determinants of Research Productivity in Drug Discovery.
NBER Working Paper. (Later published in Rand J Econ, 1996)
Originally about firm size (economy of scope) and efficiency in
management of research.
But also found,
1) Variance in research productivity explained by firm fixed effects
after controlling for firm size, scope, program size, etc.
2) Despite the fact that differences in the structure of the research
portfolio (e.g., industry) had significant effects on research
productivity, variations in portfolio structure across firms were
persistent (e.g., firm specific competence).
Guided by the resource-based view lens, the authors attempt to
explain these finings.
Why (Pharmaceutical) Research?
Successful research efforts take many years to build and often
rely on idiosyncratic search routines that may be difficult to
transfer across organizations (Nelson, 1991).
More Generally, Knowledge as exemplar of valuable, rare,
imperfectly imitable, non-substitutable resource.
In particular,
Absorptive Capacity (Cohen & Levinthal, 1990)
Combinative Capability (Kogut & Zander, 1992)
Dynamic Capability (Teece, Pisano & Shuen, 1997)
.
.
.
Architectural Competence (Henderson & Cockburn, 1994)
A Proliferation of Concepts?
Architectural Competence
Component Competence
- Abilities or knowledge specific to particular local
activities within the firm (e.g., expertise in particular
disciplinary areas)
- Fundamental to day-to-day problem solving
Architectural Competence
- Ability to use component competence
- Integrate them effectively and to develop fresh
component competences as they are required
Hypotheses
Component Competence
Hypothesis1: Drug discovery productivity is an increasing function of
firm specific expertise in particular disciplinary areas (e.g., molecular
biology, physiology, biochemistry, etc.)
Hypothesis2:
in particular disease areas (e.g., diabetics, anxiety)
Architectural Competence
Hypothesis3: Firms with the ability to encourage and maintain an
extensive flow of information across the boundaries of the firm will
have significantly more productive drug discovery efforts, all other
things equal.
Hypothesis4: Firms that encourage and maintain an extensive flow of
information across the boundaries between scientific disciplines and
therapeutic classes within the firm will have significantly more
productive drug discovery efforts, all other things equal.
Sample and Data, Measures of Variables
Population: Firms in pharmaceutical industry
Sample: 10 major European and American firms
- Data collection at research project level
- 3210 observations (research project * year)
- Data source: archival for drug discovery productivity; internal for
R&D input; qualitative for Architectural Competence
Variables
- Productivity: Counts of “Important” patent grants. (Important, if
granted in at least two out of America, EU, Japan)
- Size (R&D input): Annual expenditures on exploratory research and
clinical development by research program
- Other controls: shape, scope of research portfolio, internal and
external spillovers, therapeutic class dummies
Measures of Variables
Variables
- Component Competence
1) firm specific expertise in particular disciplinary areas: not
measured, not tested
2) firm specific expertise in particular disease areas: the stock of
patents obtained in each program (stock calculated by
assuming 20% depreciation rate for knowledge)
- Architectural Competence:
3) Information flow between the boundaries of firms: the degree
to which reputation in wider community is the dominant
criterion for promotion of scientific personnel
4) Information flow within firms (between scientific disciplines
and therapeutic classes): the degree of exchange of rich
information; the degree of regional integration; the degree of
centralized resource allocation (inverse)
Model Specification
Poisson Process
1) Poisson Distribution (mean=variance for d.v.)
if violated, estimators consistent but standard errors underestimated
2) Negative Binominal (gamma distribution for residual term)
if violated, estimators inconsistent
3) Nonlinear least-square estimates
4) quasi-generalized pseudo-maximum likelihood (or weighted 3))
For robustness check
Assumption check: ANOVA
Most of the variance for Architectural Competence variables are
accounted by between firms rather than within firms
Results
Results