Holmstrom Roberts (1999) - The Boundaries Of The Firm Revisited

From edegan.com
Jump to navigation Jump to search
Has bibtex key
Has article title The Boundaries Of The Firm Revisited
Has author Holmstrom Roberts
Has year 1999
In journal
In volume
In number
Has pages
Has publisher
© edegan.com, 2016


  • Holmstrom, Bengt and John Roberts (1999), "The Boundaries Of The Firm Revisited," Journal of Economic Perspectives, Vol. 12(4), Pages 73-94 pdf
  title={The boundaries of the firm revisited},
  author={Holmstr{\"o}m, Bengt and Roberts, John},
  journal={The Journal of Economic Perspectives},


Why do firms exist? What is their function, and what determines their scope? These remain the central questions in the economics of organization. They are also central questions for business executives and corporate strategists. The worldwide volume of corporate mergers and acquisitions exceeded $1.6 trillion in 1997. It is hard to imagine that so much time, effort and investment bankers' fees would be spent on adjusting firm boundaries unless there was some underlying economic gain. Indeed, the exceptional levels of merger and acquisition activity over the past two decades are a strong indication that economically significant forces do determine organizational boundaries.

Quick Summary

This is a beautifully written (probably by Roberts) summary/discussion paper on the boundaries of the firm. It provides:

  • An historic summary of the development of certain branches of the theory of the firm literature
  • A number of stylized facts from case studies and observation
  • A discussion of some factors, other than asset ownership, which determine firm boundaries
  • Excellent questions for future research, and many useful construct definitions

Construct Definitions

Supermodular is defined as:

"Supermodularity of a function means that an increase in one argument increases 
the incremental return from all the other arguments. 
With differentiable functions, the cross-partials are all non-negative. 
In the Hart-Moore model, supermodularity refers both to human capital and to assets,
so that having more assets implies a higher marginal return to all investments. 
See Milgrom and Roberts (1994)."

Quasi-rents gets a nice implicit definition:

"This investment is relation-specific; that is, its value is appreciably lower 
(perhaps zero) in any use other than supporting the transaction between the two parties.
Thus, once the investment has been sunk, it generates quasi-rents 
- amounts in excess of the return necessary to keep the invested assets in their current use.
There could, but need not, be pure rents returns in excess of those needed 
to cause the investment to be made in the first place."

Ex-ante and ex-post are also discussed:

"The terms ex ante and ex post-"before the fact" and "after the fact" -are widely used in this literature.
In the hold-up story, the investment must be made ex ante, before a binding agreement is reached, while
the renegotiation is ex post, after the investment. More generally, the literature refers to negotiations
that occur after some irreversible act, including the establishing of the relationship, as ex post bargaining."

Historic Literature Summary

The theory of the firm starts with [Coase (1937) - The Nature Of The Firm |Coase (1937)] - who proposed transaction costs in a world of imperfect information as the answer. [Williamson (1979) - Transaction Cost Economics | Williamson] (1975,1985, etc), refines this to center on the hold-up problem. The classic hold-up story is in Klein, Crawford, Alchian (1978) (car bodies), and was modelled by Grout (1984). It uses the notion of relational specific investments.

Property rights theory, starting with [Grossman Hart (1986) - The Costs And Benefits Of Ownership |Grossman and Hart (1996)] and refined by Hart and Moore (1990) was a distinct but somewhat similar branch.

Williamson's [Williamson (1979) - Transaction Cost Economics |TCE] paradigm has three features:

  • The Fundamental Transformation that "occurs when an exchange relationship moves from an ex ante competitive situation, with large numbers of potential trading partners, to an ex post, small-numbers one, once commitments have been made".
  • Asset-specificity is the aggregate level of quasi-rents. If this is measured by V - (V_b + V_s), only the sum (V_b + V_s), rather than the individual values matter
  • Market trade is a default that is assumed superior to within-organization trade unless levels of uncertainty, frequency, and asset specificity are high enough.

These stand in contrast to the features of the property-rights model(s):

  • Non-human assets are the defining property of the firm (the firm is its set of assets, and joint ownership implies integration).
  • Assets are bargaining tools, but unlike TCE all bargaining (including that after investments are made) is efficient - what matters is how ownership affects initial investment, which is not contractible.
  • Asset-specificity is driven by the outside option value of assets - that is their value on the market. Both of the individual values matter.
  • In general the payoffs to parties under cooperation are strictly higher than if each party goes his own way (i.e. there are complementarities), and the payoff functions are supermodular.
  • Asset control (ownership) gives incentives, and this results in a trade-off (as more assets are given to one party, less are given to another). This trade-off defines efficient ownership.
  • There is no uncertainty in these models
  • There is no frequency of transactions - this is introduced by Baker, Gibbons & Murphy (1997).
  • The level of asset specificity has no influence - investments are driven by marginal returns (which are empirically hard to observe).
  • It simultaneously addresses the costs and benefits of ownership.
  • The market gives the right to bargain and the right to exit with assets owned - it is responsible for the incentives.

The biggest flaw is in the interpretation of the identities of the parties. Why are they firms and not individuals?

The paper suggests that a true theory of the firm requires additional incentive considerations:

"...the function of firms cannot be properly understood without considering additional
incentive instruments that can serve as substitutes for outright ownership. 
Employees, for instance, typically own no assets, yet they often do work quite effectively.
In these theories asset ownership gives access to many incentive instruments and the role
of the firm is to coordinate the use of all of them. That may also explain why 
non-investing parties, including the firm itself, own assets."

Stylized Facts

The following is a short list of stylized facts from the paper, that a full theory of the firm should take into account.

Japanese firms:

  • Japanese firms have long-term, close relationships with a limited number of suppliers. The long-term relationships appear to subsubsitute for asset ownership (there is no hold-up!).
  • Asset specific investments are frequently made by Japanese suppliers.
  • Explicit contracting does not appear to be used. Transactions are repeated and reputation appears important. A small number of suppliers facilitates this by increasing transaction repetitions and allows more careful monitoring.
  • There is rich information sharing in their supplier-buyer relationships.


  • Nucor, the most successful steel maker in the US, uses an exclusive supply contract with a single firm. Despite this, hold-up does not ruin their relationship.


  • Perhaps anti-trust or tax considerations prevent mergers, but nonetheless airlines have succeeded in working together to take advantages of economies of scope.

Contractual assets:

  • Many firms (BSkyB is given as an example) own practically no assets, and yet thrive. Furthermore, other firms, like microsoft surely can not have their market value based solely on their own assets. "Contractual Assets", that is control over assets by contract (contracts that give decision rights like ownership), for instance exclusive dealing contracts, are likely the explanation for the discrepancy.

Factors that Influence Firm Boundaries

Aside from asset ownership the paper considers the following other reasons that might affect firm boundaries:

Resolving Agency Problems

Should sales force members be hired or contracted? While asset specificity might provide some guidance, it turns out empirically that measurement and agency costs are central. Employees are used when:

  • Individual performance is difficult to measure (Moral Hazard)
  • Non-selling activities are important to firms (Multi-task)

This was also born out empirically in a study of franchising comparing fast food restaurants and supermarkets, as well as in gasoline retailing. Monitoring was more important than TCE considerations.

Market Monitoring

Publicly traded firms have another (noisy but informative) measure available for performance contracting. The paper briefly considers Thermo Electron Corporation, which is an incubator. It finances and then spins-off startups. Spinoffs limit the amount of intervention T.E.C. can have in its startups post spin-off - effectively providing them with a commitment device.

Knowledge Transfers and Common Assets

"Information and knowledge problems are at the heart of organizational design,
because they result in contractual and incentive problems that challenge both markets and firms."

Arrow's paradox prevents the transfer of some information across firm boundaries. This is not a problem if the information is given away for free, and repeated interactions may help, but it still remains enormously problematic. The paper contests that knowledge transfers are a "common driver of mergers and acquisitions". Furthermore it says, "the problem with knowledge transfers can be viewed as part of the more general problem of free-riding when independent parties share a common asset. If bargaining is costly, the situation is most easily solved by making a single party responsible for the benefits as well as the costs of using the asset. Brand-names are another example of common assets that typically need to be controlled by a single entity."

Future Research

The paper says:

"It seems to us that the theory of the firm, and especially work on what determines
the boundaries of the firm, has become too narrowly focused on the holdup problem
and the role of asset specificity.  Think of arraying the set of
coordination and motivation problems that the firm solves along one dimension of a
matrix, and the set of instruments it has available along the other. Put the provision
of investment incentives in column one and ownership-defined boundaries in row
one. Let an element of the matrix be positive if the corresponding instrument is
used to solve the corresponding problem, and zero otherwise. So there is certainly
a positive entry in row one, column one: ownership does affect incentives for investment.
We have argued, however, that both the first column and the first row
have many other positive elements; ownership boundaries serve many purposes and
investment incentives are provided in many ways. 

We have argued, however, that both the first column and the first row have many other
positive elements; ownership boundaries serve many purposes and investment incentives
are provided in many ways."

It claims that many of the hybrid organizations that are emerging are characterized by high degrees of uncertainty, frequency and asset specificity, yet they do not lead to integration. Rather, high degrees of frequency and mutual dependency seem to support ongoing cooperation across firm boundaries. Apparently, Halonen (1994) may have made a start.

Finally, the paper finishes with:

"We do not believe that a theory of the firm that ignores contracts and other substitutes
for ownership will prove useful for empirical studies. The world is replete
with alternative instruments and, as always, the economically interesting action is
at the margin of these substitutions."