CoChat
Sign Up
Cited by 953Innovation and Knowledge ManagementOrganizational Management and LeadershipManagement and Organizational Studies

The DOI matches its Crossref record; 54 of the 54 references it cites were found in a registry or an index.

Temporarily Divide to Conquer: Centralized, Decentralized, and Reintegrated Organizational Approaches to Exploration and Adaptation

Siggelkow, Nicolaj, Levinthal, Daniel A.·2003·Organization Science
DOI 10.1287/orsc.14.6.650.24870
Open this paper in CoChat — free account

Abstract

To create a competitive advantage, firms need to find activity configurations that are not only internally consistent, but also appropriate given the firm's current environment. This challenge is particularly acute after firms have experienced an environmental change that has shifted the existing competitive landscape and created new, high-performing sets of activity choices. How should firms organize to explore and search such an altered performance landscape? While it has been noted that adaptive entities need to maintain a balance of exploration and exploitation, little is known about how different organizational structures moderate this balance. With the help of an agentbased simulation model, we study the value of three different organizational structures: a centralized organization, in which decisions are made only at the level of the firm as a whole; a decentralized organization, in which decisions are made independently in two divisions; and a temporarily decentralized firm, which starts out with a decentralized structure and later reintegrates. We find that if interactions among a firm's activities are pervasive, neither the centralized nor the permanently decentralized organizational structure leads to high performance. In this case, temporary decentralization—an organizational structure that has not found much attention in the literature—yields the highest long-term performance. This organizational structure allows the firm both to avoid low-performing activity configurations and to eventually coordinate across its divisions. Thus, even if the decision problem a .rm faces is not fully decomposable, a temporary bifurcation can lead to a higher long-term performance outcome. Initial decentralized exploration is, however, costly in the short run, as compared to centralized exploration. As a result, a tradeoff exists between the short-term costs of decentralized exploration and the longterm benefits of reaching higher performance. As interactions across and within divisions increase, the optimal length of decentralized exploration tends to grow. Paralleling our first result, we further show that even if a decision problem is decomposable, that is, can be perfectly modularized, it can be beneficial to create a temporary decision allocation that creates "unnecessary" interdependencies across the subsystems. This benefit arises in particular when the modules are complex by themselves. In both cases, an initial phase of exploration, enabled by an appropriate organizational structure, followed by refinement and coordination, enabled by a different structure, leads to high performance. To illustrate our general model, we focus on incumbent firms' responses to the Internet and discuss implications for the product design process.

Overview

Integrity signals

Registry

Crossref · 10.1287/orsc.14.6.650.24840

Cites

54 of 54 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

Citations

Cites

Show all 36 references

Research works printed but not linked: 10

  1. Bartlett C. A., 1990, Managing the Global Firm
  2. David P. A., 1990, Amer. Econom. Rev., 80, 355
  3. Eisenhardt K. M., 1999, Harvard Bus. Rev., 77, 72
  4. Ethiraj S., Management Sci.
  5. Khermouch G., 2001, Bus. Week, 73, 42
  6. Milgrom P. R., 1990, Amer. Econom. Rev., 80, 511
  7. Porter M. E., 1996, Harvard Bus. Rev., 74, 61
  8. 10.2307/3069394 · DOI
  9. Simon H. A., 1962, Proc. Amer. Philos. Soc., 106, 467
  10. Thompson J. D., 1967, Organizations in Action

Cited by

Sorted by citations

Items 201–300 of 950

Related

See all

Metadata

Authors
Siggelkow, N., & Levinthal, D. A.
Affiliations
University of Pennsylvania (US)
ORCID
Venue
Organization Science
Publisher
Institute for Operations Research and the Management Sciences (INFORMS)
Published
2003-12
Volume
14
Issue
6
Pages
650-669
ISSN
  • 1047-7039 (ISSN-L)
  • 1526-5455
DOI
10.1287/orsc.14.6.650.24870
OpenAlex
W4237645591
Cited by
953
References
36
Citations by year
2026: 27 · 2025: 32 · 2024: 32 · 2023: 57 · 2022: 52 · 2021: 47 · 2020: 50 · 2019: 57 · 2018: 67 · 2017: 52 · 2016: 44 · 2015: 36 · 2014: 63 · 2013: 61 · 2012: 55
FWCI
52.92
Type
journal article
Language
en
Open access
closed
Locations
Organization Science (publishedVersion)
Topics
  • Social Sciences › Business, Management and Accounting › Strategy and Management › Innovation and Knowledge Management
  • Social Sciences › Business, Management and Accounting › Management of Technology and Innovation › Organizational Management and Leadership
  • Social Sciences › Business, Management and Accounting › Organizational Behavior and Human Resource Management › Management and Organizational Studies
Keywords
organizational structure, exploration and exploitation, organizational adaptation, environmental change, modularization, interdependence, incumbent firms, internet, product design process, agent-based modeling

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review