Matching of resources and the design of organisations for project management

Abstract

This paper addresses a problem common to many high-technology firms. How can firms balance the needs of management as well as technologists in resourcing projects when both material and knowledge are needed for project success? The union between material and engineering knowledge in the form of projects within a firm is viewed as a matching, and the established literature on matching is brought to bear on the issue. In particular, the project resourcing problem is related to the well-known 'problem of stable marriage'. A formal argument demonstrates that a range of potential choices for project resourcing exists and that such choices are anchored and routinised by the different organisational designs of the firm. Efficient outcomes of project resource matching are related to several well-known organisational forms including the functional form, the project matrix, and the project-based organisation. The paper reveals that the project matrix organisation pays a heavy cost for its compromise between resource-preferable and knowledge-preferable projects.

Full text

Third International Engineering Systems Symposium
CESUN 2012, Delft University of Technology, 18-20 June 2012
Matching of Resources and the Design of Organizations
for Project Management
Scott W. Cunningham
Policy Analysis Section, Faculty of Technology, Policy and Management, Delft University
of Technology, Jaffalaan 5, 2600 GA, Delft, The Netherlands
S.Cunningham@tudelft.nl
Abstract. This paper addresses a problem common to many high-technology firms.
How can firms select and resource appropriate projects while balancing the needs of
management as well as technologists? We argue that such problems result in social
dilemmas for organizations, requiring practical institutional designs for management
and resolution. We relate this project resourcing problem to the well-known “problem
of stable marriage.” Both material as well as knowledge-based resources are needed
for project success. The formalization of the problem affords us to borrow existing
theorems from the literature on matching. We conclude that the potential for social
dilemmas inside the firm are fairly severe, directing institutional solutions into a
relatively narrow set of choices for project selection and resourcing. We relate
efficient outcomes of project resource matching to several well-known organizational
forms including the functional form, the project matrix, and the project-based
organization. The paper concludes that, while efficient, the project matrix
organization pays a heavy cost for its compromise.
Keywords. Organizational design, project management, stable matching, resource-
based perspectives of the firm
1 Introduction
The problem with which we shall be concerned relates to the following typical
situation: A firm is considering developing k projects. A successful project requires a
matching of i managers, who are gatekeepers of the resources of the firm, with j
engineers whom provide the technical know-how to carry the project to a successful
completion. The project permits a venue whereby organizational resources meet
engineering requirements. A fundamental problem of this scheme becomes the
appropriate resourcing of the respective projects given finite resources. And of course,
the resource decision is subject to strategic uncertainty whereby actors inside the firm
lobby for additional resources under a claim of need. This problem has been
addressed by Williamson (1967) in a classic study of transaction cost economics. A
further problem involves developing a consensus regarding the placement of the

engineers. Gale and Shapley (1962) address the mathematical fundaments of what has
now become known as the “matching problem.”
We contend that the difficulties here described can be addressed by the appropriate
selection of institutional form. The paper outlines a set of design variables, the
standard choices made for organizational form and project management. We relate
these choices to a class of matching problems (Gale and Shapley 1962). We conclude
by considering the adequacy of these organizational forms given various managerial
or technical imperatives.
2 Analysis
Gale and Shapley (1962) set forth the basics of the matching problem. The original
case concerned the matching of students to university in university admissions. They
also considered the matching of men and women, for instance in marriage. We now
define the setting more formally (c.f. Shoam and Leyton-Brown 2009). We further
elaborate the formalism in the context of engineering project management.
2.1 The Stable Matching Problem
These and subsequent authors model the stable matching problem with a graph G,
where the agents are represented by vertices, and two vertices are linked by an edge if
the agents are both acceptable to each other. Let M be a set of managers, let E be a set
of engineers, and let P be a set of projects. We need not assume that ∣M∣ =∣E∣=∣P∣;
thus there can be a pool of managerial or engineering talent which is currently
underemployed in the firm. Likewise there may be more projects than there are
managerial or engineering resources. We begin with the assumption that there is a
one-to-one relationship between manager, engineer and project.
At this point, we must make some stylized assumptions regarding the
organizational preferences of managers and engineers. For the purposes of
discussion, let’s examine the idea that managers are primarily motivated by resource-
based concerns. Some authors argue that management is largely about the effective
management of such resources. In contrast, let engineers be primarily motivated by
knowledge and know-how in selecting projects. There may be little overlap between
material resources and engineering know-how; the institutional problem is made
worse by knowledge asymmetries inherent in engineering projects.
Every agent s (both engineer and manager) has their own preference ordering ≻s
over projects. These orderings are likely to be different within agents. The
organizational origination and structuring of these preferences is not discussed here,
although this is a classic issue of institutional economics (c.f. Williamson 1967). With
regard to these preference orderings we write a ≻i a' to mean that manager i prefers
project a to project a'. We further use the null set ∅ to indicate when a manager or
engineer would prefer to be assigned no project at all rather than receive assignment
to project a. This is to say that project a is unacceptable to i, or ∅ ≻i a. We require the
preferences to be strict; that is to say all agents must be able to choose a preferred

option when given any two projects. Of course if a project is unacceptable, than all
projects which are unacceptable are equally unacceptable.
DEFINITION 2.1 (Project Matching). An assignment of managers or
engineers to a project is called a project matching. The project matching is
such that each manager is assigned one or fewer projects. Likewise all
engineers are assigned one or fewer projects. Let S be the set of
engineers and managers. The following are requirements of a project
matching, p(S): S ∪ P → S ∪ P ∪ {∅}. The matching is reflexive. Staff are
mapped to project, and project are mapped to staff: psp if and only if
pps. All staff members must be matched to a project ∀s ∈ S, such that
either ∃p ∈ P and thus  psp, or that staff member is underutilized
s∅.
A project matching also implies a matching between engineers and managers.
DEFINITION 2.2 (Matching). An assignment of managers to engineers via
projects is called a matching. The project matching is such that each
manager is assigned one or fewer projects. Likewise all engineers are
assigned one or fewer projects. The following matching occurs because
of project matching (M): M ∪ P → M ∪ P ∪ {∅}. The matching is
reflexive. Managers are mapped to engineers, and engineers are mapped
to managers: me if and only if em. All managers must either be
matched with an engineer, or be underutilized ∀m ∈ M, either ∃e ∈ E such
that me, m∅. The same is reflexively true of engineers.
For completeness we consider the role of under-resourced projects.
DEFINITION 2.3 (Matching Projects). We define a matching project p to
be those projects with both a manager and engineer. Matched projects q are
therefore those projects such that q(m)=q and q(e)=q. All projects in the set
P are either matched q or unmatched ¬ q. We assume that unmatched projects
map to the null set ¬ q →{∅}, and any corresponding staff is underutilized.
This definition permits us to use the three-fold distinction of engineers,
managers and projects, with a generalized two-mode graph framework
used in matching problems.
2.2 Organizational Design Choices
We consider a range of engineering project organizations as stated in the literature
(Figure 1). The prospective designs range from the fully functional form, whereby
management is allocated according to traditional functions of the firm such as
engineering, finance, or human resources. Engineering may be further divided into
separate divisions according to product lines. This is illustrated in 1A.

Third International Engineering Systems Symposium
CESUN 2012, Delft University of Technology, 18-20 June 2012
Fig 1. Design choices for functional and project-based organizations. Figure is adapted from Hobday (2000).

Third International Engineering Systems Symposium
CESUN 2012, Delft University of Technology, 18-20 June 2012
In the intermediate project matrix organization (c.f. Galbraith 1973), every project
has full commitment of management resources across all functions of the firm.
Further every project has a dedicated staff of engineers who work to provide the
project with suitable interfaces with other projects, and to provide suitable design
iterations to meet specifications. This is illustrated in 1C. Team-members in the
project matrix organization face dual responsibilities to both their project as well as
functional organizations.
The fully project-led organization consists of vertically integrated teams of
management and engineering and other staff dedicated to the needs of the specific
project. The resultant organization gains considerable flexibility for the needs of the
specific project, but abandons the synergies which can occur across divisions of the
company. This is illustrated in 1E.
In short, organizations can be arrayed in a continuum from functional to project-led
organization. There are also intermediate organizational forms, of which we choose
two of those known to the literature for further discussion. A mixed matrix-functional
form is known as a balanced matrix. A mixed matrix-project form is known as a
project-led organization. The intermediate forms have somewhat of a mixed
reputation among organizational designers as they are believed to create greater role
ambiguity and conflict among staff members (Butler 1973).
1.1 Organizations as Institutions for Stable Matchmaking
Our claim in this paper is organizational designs provide effective institutions for
the matching of resources in engineering projects. Allocation of the resources of
labor, capital, and capital goods are the typical prerogatives of managers. Equally
important for many organizations is the allocation of knowledge – both formal
knowledge, as well as learning by doing. These resources are the prerogative of
engineering staff. A failure to successfully allocate either material resources or
knowledge resources could easily gutter a project.
Institutional theorists see institutions as rules, routines or procedures for the
practical resolution of problems. This definition is widely adopted – see for instance
Hemke and Levitsky (2004). Technical problems abound in many engineering project
firms. Organizations also succumb to social problems when individual decisions are
not mediated by collective rules or procedures. Christensen (1997) describes the
“innovator’s dilemma” many high technology companies face -- yet organizations
routinely face other social dilemmas as well. Among the most severe of these
problems is surely the appropriate resourcing of internal projects.
The functional organization is a social technology which delivers the unparalleled
ability to administer resources in large organizations while holding down opportunism
and rent-seeking behaviors. Linstone (2002) argues that the transition from owner
operated firms to corporate hierarchies resulted in a dramatic increase in societal
wealth, further fueling major waves of economic activity. The functional organization
then is an institution furthering the need of owner-operators, and later shareholders
and their corporate delegates, to match resources with wealth creation opportunities

within the firm. Sosa et al. (2004) describe how traditional organizational forms
discourage needed interaction at design interfaces.
What then about the project-based organization? Hobday (2000) and others are
quite clear about this. These are the organizations excel at mustering technical
knowledge in pursuit of specialized knowledge, or customized product design.
Project-based organizations avail projects of the best available technological
organization. Thus project-based organizations are institutional solutions to the
dilemma of inadequate application of technological knowledge within the firm.
We hypothesize that the functional form of organization affords managers an
opportunity to guide the matching process between firm resources and available
projects. We further hypothesize that the project-based organization provides the
engineer with the opportunity to match their own technical knowledge with firm
projects. The mixed forms of organization shown in figure 1 are therefore mixed
responsibility organizations which variously allocate the responsibility for defining
and resourcing projects between management and engineers. This alignment of
institutional form with resource dilemmas raises still further questions concerning the
effective resourcing of projects internal to the firm.
2.3 Desirable Institutional Solutions for Matching
Three desirable qualities of a match have been identified by the literature. These
qualities can be readily applied to the resource matching problem outlined in this
section. The qualities of an effective match are rationality, lack of blocking, and
stability.
DEFINITION 2.4 (Individual rationality). A matching is individually
rational if no staff s prefers to remain unmatched than to be matched to (s).
DEFINITION 2.5 (Lack of Blocking). A matching is unblocked if there
exists no pair (m,e) such that (m) ≠ e, but e ≻m (m) and m ≻e (e).
Definitions 2.4 and 2.5 can be intuitively understood as the following. A solution
is well-matched when no staff member is matched to an unacceptable project, and
when no pair of managers and engineers would be preferred to be matched to each
other than their respective projects. Putting these together we have the definition of a
stable matching.
DEFINITION 2.6 (Stable Matching). A matching is stable if and only if it is
individually rational and unblocked.
Given this problem definition, subsequent structuring and definitions, we are now
prepared to advance hypotheses concerning the role of an effective organizational
design in resourcing projects. In section 3 we build upon existing theorems of
matching. In addition we introduce some new corollaries to existing findings, with
import for project management. The nature of project “resource buckets” and the
likely response of organizations to environmental uncertainty has been considered by

Chao and Kavadias (2008). Project stability and resource appropriation is a matter for
serious concern.
3 Organizational Design and the Effective Resourcing of Projects
The following section used a mixed form of reasoning, switching between the
applied problem context of project resources, and the formal domain of matching
problems. Nonetheless we hope to make it clear how abstract findings relate to the
application setting. Our first finding is the conclusion that the matching of resources
to projects in organizations is in fact a social dilemma, which cannot be solved using
solely informal methods.
THEOREM 3.1. If there is two-sided opportunism or strategic behavior, the
matching problem cannot be solved in absence of a formal institution.
Proof (Roth 1982). The proof consists of a proof by contradiction showing
that individual agents have no incentive to truthfully reveal their preferences
for projects given a lack of knowledge of the preferences of others on the
staff. If they do truthfully reveal such preferences then this will be exploited
by other staff. Furthermore, there is no effective project selection mechanism
which can occur in absence of knowing the preferences of others (ex post
equilibrium).
It is possible to solve the problem if either engineers or managers preferences are
common knowledge. This could occur for instance if managers were in a contractual
structure to manage resources and effort. This is a common assumption concerning
the governance of hierarchies. Such a contractual mechanism would in fact constitute
a formal institution. A second conclusion, building upon the literature, is that if you
were to design an project organization, it is in fact possible to find a stable matching
between engineers and managers using projects.
THEOREM 3.2. There exist organizational designs which can provide an
effective matching of projects to resources.
A further conclusion is that there may be many such matchings; we may seek further
for only those designs which are optimal for either engineer or management.
DEFINITION 3.3. A stable matching  is optimal if every staff member likes it
at least as well as any other stable matching; that is, ∀s ∈ S and for every
other stable matching ''s≽s s

Recall our hypothesis of section 2.3 that functional forms of organization afford
managers their optimal choice of projects. Likewise, consider the hypothesis that
project-based organizations afford engineers their optimal choice of projects. The next
theorem demonstrates that such institutional forms do exist, and furthermore there is
only one such institution possible.

THEOREM 3.4. There is an institution, which we call the “functional
organization” which is capable of achieving a stable matching which
furthermore is optimal for managers.
THEOREM 3.5 (Roth and Vande Vate 1990). Roth and Vande Vate suggest a
proposal rejection algorithm which permits projects to be proposed in any
order. Stability can be achieved by a proposal-rejection process.
Thus institutions capable of effective matching of diverse resources within an
organization are definitely possible. Such an institution can also manage the
decentralized creating and ranking of projects. The institution can work even as new
projects are proposed and are entered into consideration for resource sharing. An
optimal institution is as equally feasible for engineers as it is for managers. This leads
to corollary 3.5.
COROLLARY 3.5. There is an institution, which we call the “project-based
organization” which is capable of achieving a stable matching which is
engineering optimal.
THEOREM 3.6. There exists only one engineering optimal and only one
management optimal institution.
Proof. The Gale and Shapley algorithm can be run by either party, resulting
in two distinct but singular and stable solutions.
A logical counter-argument would be to ask whether or not a single institution
could be both engineering optimal and management optimal at the same time. Here
again we have existing proofs from the literature that engineering optimal and
managerial optimal solutions are frequently mutually incompatible.
THEOREM 3.6. Any organizational form which is stable and which is optimal
from a managerial perspective will necessarily be the worst achievable from
an engineering perspective. The reverse is also true.
Proof. An optimum matching is by definition Pareto optimal. If the managers
choose the projects, then they will necessarily choose the projects which are
best for them, requiring engineers to accept second-best projects. The obverse
will be true if engineers are in the lead. The asymmetry of outcome is caused
by either managers or engineers taking the first choice of projects.
We are then lead to question whether intermediate forms of organization, such as
the project matrix, make sense given a project context.
THEOREM 3.7. Intermediate forms of organization (project matrix) are
possible, but result in all projects being second-best from either an
engineering or a managerial perspective.

Proof. The proof follows directly from stability concerns. For there to be a
stable solution, then all managers must be better off with their current project
than with any other project available to them. The same must hold for
engineers. Furthermore, if the institution is indeed intermediate between
engineering and management, then neither member of staff can be given their
first-best choices. Such an organization may be implemented in a
decentralized fashion using a project rejection scheme, preventing either
engineering or management from dominating the process (Roth and Vande
Vate 1990).
Thus, the matrix organization enforces a standard of mediocrity across all projects
in the organization. Such uniformity may be still be desirable when mutual losses for
management and engineering staff are high for substandard projects.
4 Concluding Remarks
A number of areas for future research are suggested. First, this line of
argumentation could provide pay-offs in examining project selection procedures. The
existing results concerning appropriate project deferral and rejection might be used in
existing project management organizations in pursuit of efficient project selection and
funding. Furthermore in the framework considered here, projects are a “meeting
point” for resource exchange engineers and managers. A more general framework
could permit weighted assignments, given the so-called “stable allocation problem
(Biró 2007). ” This raises a larger issue concerning the appropriate number of
projects, as there is clearly more to the problem than meeting resource -based
constraints.
Practical relevance also requires we consider the dynamics of project design and
selection. Can an optimal allocation of projects be maintained as the project portfolio
grows or changes? Similarly, if organizational designers select a given institutional
design, how can they be sure that the design isn’t subverted by the awarding of
projects? The institutional setting is need not be fixed. Surely one of the more
prominent modes of failure of the project matrix organization is when a functional
organization permanently acquires resources only temporarily allocated during project
resourcing.
As Gale and Shapley (1962) concluded before us “in making the special
assumptions needed in order to analyse our problem mathematically, we necessarily
moved further away from the original . . . question, and eventually . . . entered the
world of mathematical make-believe (p. 15).” Nonetheless we think “the world of
mathematical make-believe” has value for considerations of organizational design and
project management, if only to remind us of the severity of the choices facing the
firm. No informal mechanism exists for project selection, and any formal institutional
mechanism must make difficult trade-offs between resources and technical
knowledge. Any admixture of both resources and technical knowledge must
necessarily result in a hybrid organization adequate at many criteria, but be excellent
for very few.

References
Arrow, W. B. (2009), The Nature of Technology: What it is and How it Evolves. Allen Lane:
London.
Biró, P. (2007), The Stable Matching Problem and its Generalizations: An Algorithmic and
Game Theoretical Approach, PhD Thesis. Budapest University of Technology and
Economics: Budapest.
Butler, A. (1973), Project management: A study of organizational conflict. Academy of
Management Journal, 16: 84-102.
Chao, R.O., Kavadias, S. (2008), A theoretical framework for managing the new product
development portfolio: When and how to use strategic buckets, Management Science, 54(5):
907-921.
Christenson, C. (1997), The Innovator’s Dilemma: The Revolutionary Book that will Change
the Way You Do Business, Harvard Business Review Press: Harvard, MA.
Galbraith, J.R. (1973), Designing Complex Organizations. Addison-Wesley: Reading, MA.
Gale D., Shapley L.S. (1962), College admissions and the stability of marriage, The
American Mathematical Monthly, 69(1):9-15.
Helmke, G., Levitsky, S. (2004), Informal institutions and comparative politics: A research
agenda, Perspctives on Politcs, 2(4): 725-740.
Hobday, M. (2000), The project-based organization: An ideal form for complex products
and systems, Research Policy, 29: 871-893.
Knight, K. (1976), Matrix organization: A review. Journal of Management Studies, 13(2):
111-130.
Linstone, H. (2002), Corporate planning, forecasting, and the long wave. Futures, 34: 317-
322.
Mintzberg, H. (1980), Structure in 5’s: A Synthesis of the Research on Organizational
Design, Management Science, 26(3): 322-324.
Roth, A.E. and Vande Vate, J.H. (1990), Random paths to stability in two-sided matching.
Econometrica, 58(6): 1475-1480.
Roth, A.E. (1982), The Economics of Matching: Stability and Incentives, Mathematics of
Operations Research, 7: 617-628.
Shoam Y., Leyton-Brown, K. (2009),. Multiagent Systems: Algorithmic, Game-Theoretic,
and Logical Foundations, Cambridge University Press: Cambridge.
Sosa, M.E., Eppinger, S.D., Rowles, C.M. (2004), The misalignment of product architecture
and organizational structure in complex product development, Management Science, 50(12):
1674-1689.
Williamson, O.E. (1967), Hierarchical control and optimum firm size. Journal of Political
Ecconomy, 75, 123-138.