Isomorphism of Vertical Integration & Contracting: Theory of Vertical Control, Study notes of Literature

Download Isomorphism of Vertical Integration & Contracting: Theory of Vertical Control and more Study notes Literature in PDF only on Docsity!

BEBR

FACULTY WORKING PAPER NO. 89- College of Commerce and Business Administration University of Illinois at Urbana- Champaign December 1989 The Choice of Organizational Form: Vertical Integration versus other Methods of Vertical Control Joseph T. Mahoney Department of Business Administration University (^) of Illinois 1206 South Sixth Champaign, IL 61820 (217) 244-

The Choice of Organizational Form:

Vertical Integration^ versus^ other^ Methods^ of^ Vertical^ Control

Abstract

Vertical integration is a fundamental corporate strategy of interest to the fields of

strategic management and organizational^ economics.^ This^ paper synthesizes^ theoretical arguments

and empirical findings from^ this^ literature^ to identify^ the^ underlying^ advantages and

disadvantages of choosing a vertical integration strategy. It then suggests that these arguments

can be put into a broader framework with vertical financial ownership and vertical contracting

being seen as essentially isomorphic phenomena. The key theoretic question involves predicting when market mechanisms are sufficient, when intermediate forms of vertical contracting (^) become necessary, and when formal vertical integration is the preferred strategic (^) alternative. The

concluding section of the paper provides a framework for making this analysis based on a

synthesis of agency and transaction cost perspectives.

The Advantages of Vertical Integration An exhaustive review of the economic and strategy literature^ (Mahoney, 1989) suggests

that the motives for vertical^ integration^ may^ be^ classified^ into^ the^ four^ major^ categories^ shown^ in

Table I: (1) transaction^ cost^ considerations;^ (2) strategic^ considerations;^ (3) output^ and/or^ input

price advantages; and (4) uncertainties^ in^ costs^ and/or^ prices.

Insert Table 1 about here

While no firm will^ be motivated by all^ of^ these^ potential^ advantages,^ taken^ as a whole^ they

illustrate the broad utility of this corporate option, a usefulness that justifies greater theoretical

and empirical attention than^ has^ been^ given^ to vertical^ integration^ strategies^ to date.

Transaction Cost Considerations. Vertical integration is often chosen to minimize the cost

of transactions (Coase, 1937; Williamson,^ 1985), costs which^ include^ negotiating,^ adapting,

monitoring, and enforcing buyer-supplier relationships^ (Jensen^ & Meckling,^ 1976). A^ good example of the potential cost savings of vertical merger is the avoidance of sales taxes when arms- length contracting is replaced by internal transfers (Coase, 1937). More subtly, vertically

integrated petroleum firms have found it profitable to increase the price of crude oil relative to

the price of final products in order to shift as much of their reported earnings as possible to the

raw materials extraction stage, which enjoys tax preferences associated with resource depletion

(Bolch & Damon, 1978). Similar strategies can be found in other basic conversion industries such as copper, aluminum and steel (Scherer, (^) 1980). A second fundamental motive for vertical integration is the failure of markets to

satisfactorily handle certain transactions (Casson, 1984). Important sources of market failures

include externalities (Dahlman, (^) 1979), increasing returns and sunk costs (Baumol, Panzar (^) & Willig,

1982) and market imperfections (Wolf, 1979; Yao, 1988). These market frictions violate the

standard assumptions of competitive equilibrium models. Prices are no longer sufficient statistics.

Long-term relational contracts (Macneil, 1980), joint ventures (Harrigan, 1988; Hennart, 1988a;

Kogut, 1988), franchising (John, 1984; Norton, 1988; Rubin, 1978), networks (Thorelli, 1986;

Jarillo, 1988), quasi-firms (Eccles, 1981) and^ vertical^ financial^ ownership^ (Harrigan,^ 1983;

Williamson, 1971) are^ some^ of^ the^ "institutions^ of^ capitalism"^ (Williamson,^ 1985)^ which^ emerged

in response to the inadequacies of^ classical^ market^ contracting.

Williamson's (1985) seminal^ research^ develops^ a^ well-grounded^ theoretical^ framework^ for explaining and predicting this market failure. The basic idea^ is^ that^ contractual^ difficulties^ arise when opportunistic agents (Anderson, 1988; Maitland, Byrson & Van De Ven, 1985) engage in

frequent transactions in an environment^ of^ sufficient^ uncertainty^ and/or^ complexity^ to^ surpass

bounded rationality^ capabilities^ (Simon,^ 1978).^ The^ risk^ of^ self-interested^ agents^ utilizing

asymmetric information to their advantage is high in such environments^ and vertical financial

ownership is one response to this inadequacy of classical market contracting (Harrigan, 1983;

Williamson, 1975). Contractual problems become acute when^ there^ are^ small^ numbers bargaining, a situation that occurs when transactions involve human, physical or site "asset

specificity" (Spiller, 1985; Williamson, 1979). Human asset specificity involves uniquely related

learning processes or teamwork. Physical asset specificity includes requirements for specialized

machine tools and equipment. Site specificity occurs when unique locational advantages exist, (^) as, for example, when a power plant is located near a coal mine to save on transportation costs

(Joskow, 1985a). Vertical integration can assure requisite inputs in such situations and the

importance of asset specificity in explaining and predicting vertical financial ownership is

supported (^) by a large (^) body of literature including case studies (Butler & Carney, 1983; Globerman & Schwindt, 1986; Goldberg & Erickson, (^) 1987; Hennart 1988b; Klein, Crawford & Alchian, (^) 1978; Palay, (^) 1984; Stuckey, (^) 1983; Teece, 1976), formal modeling (Kleindorfer & Knieps, 1982; Masten, 1982; Riordan & Williamson, 1985) and statistical testing (Anderson (^) & Schmittlein, (^) 1984; Armour & Teece,^ 1980; Caves & Bradburd, 1988; John & Weitz, 1988; Joskow, 1985a; Levy, (^) 1985; MacDonald, 1985; MacMillan, Hambrick (^) & Pennings, (^) 1986; Masten, 1984; Monteverde & Teece, 1982; Walker & Weber, 1984, 1987). A last important transaction (^) cost motive for vertical integration involves economies of scope (Baumol, (^) Panzar, & Willig, 1982; Teece 1980; Williamson, 1975), including technological complementarities (Bain, (^) 1968). The standard example (^) of vertical merger to achieve economies of

difficult for^ an^ oligopolist^ to^ plan^ secretly^ to^ increase^ market^ share.^ Since^ there^ were^ few significant independent^ refiners,^ no^ company^ could^ increase^ its^ output^ of^ crude^ oil^ without^ first

building refineries and distribution systems^ which^ clearly^ signaled^ their^ plans^ to^ competitors.

Vertical integration thus^ may^ evolve^ as^ a^ means^ of^ maintaining^ oligopolistic^ discipline^ and^ may

provide mobility barriers (Caves & Porter,^ 1977) which^ sustain^ the^ stability^ of^ strategic^ groups (McGee & Thomas,^ 1986; Newman,^ 1978).^ Differences^ between^ existing^ firms^ in^ their^ degree^ of

vertical integration appear to have^ led^ to^ difficulties^ in^ agreement^ on^ the^ desirable^ vertical^ price

structure. Even more problematic, changes in vertical integration structure can^ increase^ the threat of entry as it^ has^ in^ steel^ (Adams^ &^ Dirlam,^ 1964), petroleum^ (de^ Chazeau^ &^ Kahn,^ 1959), and aluminum (Scherer, 1980).

Output and Input Price Discrepancies. If output and input prices are not given to the

firm, then there are several possible explanations^ for^ vertical^ integration.^ In^ the^ successive

monopoly case, Spengler (1950) considers a product that passes through three successive stages of

production before being ready for sale to consumers. Each stage of production contains sufficient

monopoly power to charge a price above the competitive level. Here, a vertically integrated firm

controlling all three stages of production can earn a larger profit than can be obtained by the

"myopic chain monopoly" (Greenhut & Ohta, 1976). The essential idea is that the vertically

integrated producer can evade the monopoly prices imposed by upstream firms.

In the case of bilateral monopoly (Machlup and Taber, 1960), vertical integration

facilitates arriving at the input choice consistent with joint profit maximization under non-

integrated bilateral monopoly (Williamson, 1975). Internal organization minimizes strategic

bargaining and resolves the conflict of the division of profits. The problem of firms being "locked in" (^) to a vertical (^) relationship is not uncommon (Klein, Crawford & Alchian, 1978).

Vertical integration minimizes appropriation risk (Barney, 1986; Walker, 1988).

In the case of an upstream monopoly, if there exists a variable-proportions technology

(Warren-Boulton, 1978) for the final product, vertical merger permits the integrated firm to

achieve efficiency in factor utilization. As a first approximation these cost savings accrue as

additional (^) profits to the integrating monopolist (Mallela & Nahata, 1980; McGee & Bassett, 1976;

Schmalensee, 1973; Vernon^ &^ Graham,^ 1971;^ Waterson,^ 1982).^ Abiru^ (1988)^ extends^ this^ stream of literature to include the^ more^ empirically^ relevant^ case^ of^ variable-proportions^ technology^ and successive oligopolies. The price discrimination incentives for vertical integration can be elucidated by the example of an intermediate good monopolist^ selling^ to two^ downstream^ competitive^ industries. The upstream monopolist can^ increase^ profits^ by^ selling^ the^ intermediate^ product^ at^ a^ lower^ price to the^ downstream^ firm^ with^ the^ relatively^ higher^ price^ sensitivity^ (Perry,^ 1978).^ Vertical

integration by the upstream monopolist can eliminate^ incentives^ for^ arbitrage^ of^ the intermediate

product between the downstream firms. Perry (^) (1980) contends that forward integration by Alcoa

in the period 1888-1930 was inspired by price discrimination. Alcoa integrated into the relatively

more price sensitive markets, such as cookware (Hale,^ 1967).

' Crandall (1968) submits that Ford's expansion into competitive components markets

similarly was motivated by the desire to price discriminate. Ford purchased Auto-Lites battery

plant and obtained more revenue from those who used their vehicle the most (Weintraub, 1949).

Asset specificity of parts and economies of scale in producing repair parts also tend to lock

customers in with an automobile manufacturer.

Uncertain Costs and Prices . Vertical integration is a potential response to the stochastic

elements confronting the firm. Arrow (^) (1975) examined uncertainty in the supply price of the

upstream good by focusing on asymmetric information between parties at the upstream and

downstream stages. A downstream firm has the incentive to purchase one or more upstream firms

because this improves its pricing forecast and thus its ability to purchase the appropriate level of

capital. Carlton (1979) presents a similar model in which both output and input firms face

uncertainty in demand and firms must make decisions concerning price and production before actual (^) demand is observable. In this case there is some risk of supply failure to the customer as

well as risk to the seller of overproduction. Vertical integration is a means of transferring this

risk. Firms integrate (^) to ensure a supply of input for their "high probability" demand and continue

to purchase their "low probability" demand.

induces vertical^ integration^ (Jones,^ 1984).^ When^ output^ depends^ on^ joint^ efforts,^ individuals^ have

the incentive to "free-ride" in^ hopes^ of^ receiving^ greater^ reward^ than^ their^ efforts^ would

otherwise dictate. Empirical studies are consistent with the hypothesis that measurement

uncertainty of this type leads to vertical^ integration^ (Anderson^ &^ Schmittlein,^ 1984; Anderson, 1985). Measurement uncertainty and quality^ uncertainty^ are^ also^ important^ factors^ that^ lead^ to

performance ambiguity (Jones, 1987). The need to reduce quality uncertainty for key inputs may

spur backward integration, while the need to assure point-of-sale service, which is often critical

for new products, may necessitate forward integration (Harrigan, 1986).

Finally, the problem^ of^ technological^ uncertainty^ (Hennart,^ 1982; Teece,^ 1982) and^ the

trading of technological knowledge may lead to vertical integration (Arrow, 1971). Here again an

apparent disagreement can be found in the literature. Armour^ and^ Teece^ (1980) argued that the

strong relationship between research intensity and vertical integration in the petroleum industry

was due to market failures in information exchange. However,^ Harrigan (1986) and Walker & Weber (^) (1984, 1987) found that technological uncertainty was associated with (^) less vertical integration. The resolution of apparent disagreement here requires care to not confound (^) asset specificity and uncertainty. If technological uncertainty leads to the utilization of more flexible

(less firm-specific or product-specific) technologies, a link suggested by Balakrishnan and

Wernerfelt's (1986) model, then less vertical integration obtains.

The problems of recognition, disclosure, team organization and dissipation that are

involved in contracting under technological uncertainty all suggest a decision to vertically

integrate (Caves, (^) 1982; Teece, (^) 1982). The effect of technological uncertainty on vertical

integration may be especially influenced by the coordination costs of contracting for many parts

in a system. Monteverde and Teece (1982) argued that the automobile electrical system involved

substantial (^) interdependencies and were consequently produced in-house. The Walker and Weber

(1984, 1987)^ automobile^ studies^ could be updated to consider^ these^ system^ coordination^ influences

on the technological uncertainty—vertical integration linkage.

An Overview of Vertical Integration as a Corporate Strategy Uncertainty can take many forms. (i) Parametric or^ structural^ (Langlois,^ 1984). (ii) perceptual or market based (Downey, Hellriegel, & Slocum, 1975).

(iii) volume,^ measurement,^ quality,^ or^ technological.

In the absence of any of these uncertainties, the^ firm^ need^ not^ exist^ (Coase, 1937; Knight, 1921;

Williamson, 1975). As uncertainty^ increases,^ not^ only^ is^ the^ firm^ called^ into^ existence,^ there^ are

increasing arguments for expanding the scope of organizational activity through vertical

integration. More specifically, the same arguments found in the basic theory of the firm (Coase,

1988; Penrose, 1959) can be extended^ to justify^ vertical^ integration^ as a^ corporate^ form^ (Harrigan,

1983; Williamson,^ 1975). To ascertain the^ effect^ of^ increased^ "uncertainty"^ on^ vertical^ integration,^ however,

researchers must specify the nature of the uncertainty, the context of the analysis (static or

dynamic), and the interaction of uncertainty with other important variables such as asset

specificity (Anderson, 1985; Walker & Weber, 1984). Clearly, the economic and strategy literature has offered an impressive menu of advantages for the corporate strategy of vertical

integration. But what are the advantages of vertical integration that are generalizable?

(1) Profit.^ Vertical^ integration^ may^ most^ effectively^ achieve^ the^ profit^ incentive^ since

preemptive claims on profits between separate firms are eliminated.

(2) Coordination.^ The^ firm^ has^ better^ control^ of^ opportunistic^ behavior^ due^ to the

authority relationship (Dow, 1987) within the firm. Managers of the divisions can be required to

cooperate and promotions can be adjusted to achieve such behavior.

(3) Audit^ and^ Resource^ Allocation.^ The^ auditing^ powers^ of^ the^ firm^ are^ superior^ to the auditing capabilities of contracting parties (Williamson, 1975). A firm has the legal right to audit

its divisions but no right to audit outside contractors. Integrated firms have superior information

upon which they can base allocations to their divisions so that the incentive for those divisions to

use their (^) information strategically (to the detriment of the enterprise's profits) is eliminated

(Crocker, 1983). Furthermore, improved information enables the firm to allocate personnel to

literature.

The disadvantages of^ vertical^ integration,^ summarized^ in^ Table^ 2,^ may^ be^ classified^ under

four major^ categories:^ (1) bureaucratic^ costs;^ (2)^ strategic^ costs;^ (3) production^ costs;^ and^ (4)

long-run dynamic costs.

Insert TABLE 2 about here

Bureaucratic costs. Implementation costs of integration have proved to have particularly

important negative effects, especially because they^ are^ so^ difficult^ to^ anticipate.^ Vertical^ merger

increases the size of an organization which often results in additional hierarchical levels.

Increasing size and bounded spans of control^ imply^ greater^ distance^ of^ most^ subordinates^ from

their ultimate superiors. This may lead to communication distortion due to serial reproduction

loss and/or deliberate distortion to achieve divisional objectives (Calvo & Wellisz, 1978; Coase,

1988; Cremer,^ 1980; Williamson,^ 1967) thus^ obviating^ a^ major^ advantage^ of^ integration.

The loss of high powered market incentives suggests that internal organization may also be

more costly than the market mechanism (Williamson, (^) 1985), undercutting the profit incentive for integration. One explanation (^) is that the lack of direct competitive pressures on the cost of the intermediate products may allow increasing levels of slack (Cyert & March, (^) 1963) and thus reduce (^) profitability. As firms vertically integrate away from the (^) base business, they are also likely to become involved in new manufacturing or selling tasks. Managing at the manufacturing and distribution

stages requires different skills than previously required by the firms only in upstream or

downstream operations and inexperience may lead to comparatively high internal costs (Harrigan,

1985c; Porter, 1980). In short, the synergies created through vertical integration may be

overestimated and do not compensate for higher costs (Harrigan, 1984).

Strategic costs. While Arrow (1975) suggested that vertical integration may eliminate the

problem of asymmetric information, the flip side of the argument has been suggested by Harrigan

(1984), namely^ that^ vertical^ integration^ may^ result in a loss of access to information and tacit 11

knowledge as relationships^ with^ experienced^ and^ more^ broadly^ based^ distributorships^ are^ severed. A second potential strategic cost to vertical integration is that the firm purchases specialized assets

that increase sunk^ costs and^ may^ lead^ to^ chronic^ excess^ capacity^ and^ low^ profitability^ (Baumol,

Panzar & Willig, 1982; Chandler, 1962; Rumelt, 1974). Third, vertical integration may link a firm to a weak adjacent industry^ (Harrigan,^ 1984)^ with^ an^ attendant^ loss^ in^ profitability.^ Fourth,

vertical integration may decrease a firm's strategic flexibility and lead to high exit barriers

(Harrigan, 1985d). Production costs. Walker & Weber (1984) suggest that production costs are critical in the make-or-buy decision. A vertically integrated firm that does not utilize a sufficient amount of

the input to achieve minimum efficient scale will be at a cost disadvantage against firms that

contract out to an efficient supplier achieving full economies of scale (Stigler, 1968). Second,

obsolete processes may be perpetuated with vertical integration (Harrigan, 1984). Third, vertical

integration may lead to a capital drain, a potential problem that is particularly damaging to

smaller firms (Williamson, 1975). Fourth, capacity imbalance in the vertically integrated firm

may lead to higher production costs than incurred by firms that utilize market mechanisms (Hayes

& Wheelwright, 1984). Long-run dynamic (^) costs. The make-or-buy decision also must take into account

possible long-run costs. Internal organization may lead to inefficiency if, for example, internal

divisions face no competitive pressures for procurement. Even if outside sources exist as a

potential disciplining influence, they may be bypassed due to bureaucratic considerations.

Second, (^) a norm of reciprocity (^) between divisions easily develops (Gouldner, 1960), and over time

the benefits of reducing transaction costs are lost. Third, internal production may lead to an

expansionary bias that compromises cost minimization (Williamson, 1975). Fourth, psychological

commitment and (^) administrative difficulties of divestment (Duhaime & Grant, 1984; Duhaime & Schwenk, 1985) are important dynamic (^) costs that need (^) to be considered in the make-or-buy decision. 12