does not resolve many of the problems which historically made the tax authorities fight shy of explicitly adopting a profitsplit method. Hence, it is not much help in
evaluating the profit split between the related parties: by focussing on `measurable ¶
functions, it merely assumes that synergy profits accrue to the ultimate parent. Due to their broader political and economic implications, there is likely to be
continued reluctance to tackle these questions explicitly and publicly, and a continuing insistence that the normal and primary test of transfer prices is that of
independent parties dealing at arms length. Indeed, the White Paper was fiercely criticised by the Taxation Commission of the International Chamber of Commerce, as
entailing `fundamental deviations from the arms length principle
¶ Professor Wolfgang Ritter, responding to the White Paper on behalf of that Commission, wrote
that the Basic Arms Length Return Method: is nothing but another form of unitary taxation, now recognised by the US
Federal Authorities, as a result of the international outcry which it provoked, to be unfair. BALRM analysis does exactly what unitary taxation does:
instead of looking at the actual prices charged between related entities, it looks at the total profits earned by a number of entities and divides them up
by reference to criteria determined by academic economists with no experience of the real business world. Ritter, in De Hosson 1989, p.73.
4. Transfer Prices in Theory and Practice
The detailed analysis in the previous sections has shown the difficulties faced in applying the arms length criterion. The initial adoption of the arms length principle
followed from the fear that any attempt to reach international agreement on general principles to define and allocate the tax base of internationallyintegrated businesses
would fail. Instead, it was hoped that the issues could be resolved either `naturally
¶ by using market prices, or by casebycase agreement or direct negotiation between
tax authorities. Reliance on market prices has provided at best a partial solution, principally because experience has shown that the criterion of comparable
uncontrolled prices is an inadequate one. In a high proportion of cases, the internal transactions of an integrated firm are not comparable with those of independent firms.
Furthermore, effective verification of prices by reference to comparables requires a sophisticated administrative bureaucracy, which must grow in proportion to the
importance of TNCs in the economy. Although safeharbour or brightline rules may facilitate administration e.g. the Italian rules on royalty payments, or the rulesof
thumb sometimes used for thin capitalization they have not been generally adopted: if binding they are inflexible, and if presumptive they are uncertain see section 3c
above. Significantly, it was the largest state, the US, which made the most determined efforts to operationalise the arms length criterion based on comparables,
and which has been forced to develop alternatives.
4.a CostSharing and ProfitSplit
It was, however, always accepted that the arms length principle need not and could not rely exclusively on finding comparable market prices. Its essence was that
taxation of the business profits of a branch or affiliate must begin from the separate accounts of that entity and proceed by adjusting specific transaction prices, to
produce the result which would have obtained if the entity had been independent. This entails i analysing the intrafirm relationships so as to identify and correctly
characterise specific transactions, and ii establishing criteria for pricing in the absence of market comparables which conform with the independent enterprise test.
The difficulty of identification of specific transactions which can be priced is most acute in relation to joint and fixed cost items. This has led to a trend towards the
proportionate allocation of such costs; while the problem of finding an independent enterprise criterion for pricing has led to the use of an arms length profit standard.
We have considered above the problems involved in `unbundling ¶relationships and
characterising transactions in terms of supplierpurchaser. We have also seen that pricing in the absence of comparables leads to serious problems in determining the
allocation of costs. It is very difficult to allocate the joint costs and overheads of an integrated firm according to the specific benefit derived by each unit. This has led to
the acceptance, in varying degrees, of `formula
¶methods of allocating joint costs: in particular, the averaging of interest expense required by the US and Japan, but not
accepted by others, see section 3.c.iii above; and costsharing for central services more generally accepted, but subject to strict conditions, see section 3.d above, as
well as costcontribution for research section 3.e above. Significantly, it is representatives of business who have pressed for acceptance of proportionate sharing
of costs, although they are generally hostile to a formula approach to taxation. However, while some national authorities require the proportional allocation of some
items of cost, others are unwilling to accept such an allocation without proof of specific benefit, and the principle has been rejected by a majority for some items,
notably interest expense.
Reliance on an arms length profit standard has resulted, in practice, from the problems of defining criteria for arms length prices in the absence of comparables.
Both in their administrative rules or guidelines, and as part of enforcement practice, the national authorities have relied extensively on comparison of the profits declared
by the local entity with those of similar firms normally managed. This method has been justified not as a primary means of allocating profits, but as a check on the
acceptability of price adjustments. Also, in principle it consists not in dividing the global profit of the firm as a whole, but in estimating, based on the value of assets or
the volume of business, whether the profit of the local branch or subsidiary is in line with the levels of profit made by similar independent businesses. This is similar
to the `empirical ¶ method which was given limited formal authority in the
international model treaties.
1
Even this test has led to disagreements, specifically as to whether the comparison should be with global industry standards or those of other
local firms section 1.c above. More fundamentally, the profit comparison approach entails a shift from the arms length pricing rule to an arms length profit standard,
which is controversial. Thus, when the majority of the members of the OECD committee agreed to treat thin capitalization as a transfer pricing problem, Germany
objected in principle on the grounds that the arms length principle means arms length prices and not arms length profits section 3.c.i above.
For a number of reasons, however, profitcomparison tends to become profitsplit. Inevitably, a profitcomparison will also be made between the profits made by the
local entity and those of the related companies with which it deals. Further, since profit comparison is used when there are no substantially comparable independent
party transactions, it is likely that there will be no substantially comparable firms either. In such circumstances, the experience with the US rules has shown that the
resaleminus, costplus and other alternative methods to CUP tend to result in profit split. This is because the application of this type of cost analysis to an integrated TNC
often does not result in one precise price, but a range or continuum. Applying micro economic marginal analysis tends to produce a significant gap between the marginal
price which will induce the producing affiliate to supply, and the price at which the marketing affiliate can make a profitable resale. Hence, both tax examiners and courts
tend to `split the difference
¶and adjust the price so that a `reasonable¶profit split is produced.
2
Finally, to the extent that it is difficult or impossible to define specific transactions and adjust prices on the basis of market comparables, profitsplit
becomes not a fallback but a principal method accepted by the US courts in Eli Lilly: see section 3 a above.
This is implicitly recognised by the US White Paper proposals. The White Papers Basic Arms Length Return Method attempted to establish a more precise and detailed
methodology for separate accounting in the absence of comparables, by the identification of factors of production and allocation to them of an appropriate rate of
return. This is an attempt to make a costplus and profitcomparison approach work, by arguing that as long as there is functional comparability, it does not matter if there
is no exact comparability of products. However, the authors of the White Paper were also clearly aware that this process might result in some very disproportionate profit
splits between the parent and its foreign subsidiary, since the BALRM method assumes that the additional profits from synergy are attributable to the parent. Hence,
1
OORFDWLRQE\CSHUFHQWDJHRIWXUQRYHU¶IRUEUDQFKHVZDVDFFHSWHGDVDIDOOback method: see section 1.c and Chapter 1.5.b above.
2
This point has been particularly stressed by Langbein: see Langbein 1986, p.658; Langbein 1989, p.1395.
its concession that in some cases it may be appropriate for the `residual ¶profit to be
apportioned in some appropriate manner. Unfortunately, here the methodology failed, and no criterion could be proposed on which to base this profitsplit: it must be fixed
ad hoc by agreement among the parties concerned.
The move towards acceptance of proportional allocation of joint costs and the use of an arms length profit standard are practical attempts to adapt the arms length
principle to the realities of the intrafirm relationships of an integrated TNC. While they still begin from separate accounts, they address the question of criteria for
allocation of costs and profits of the firm as a whole. However, proportional allocation of costs, as we have seen, is controversial and remains anomalous without
a global approach to profit allocation. On the other hand, profitcomparison based on return to assets or other methods may merely identify large residual profits, for the
allocation of which no criteria have been suggested.
The arms length approach attempts to maintain tax neutrality and equity in the allocation of international investment: by treating members of a group in the same
way as unrelated companies, their tax burden should be the same and markets would not be distorted by taxation. The validity of this approach, however, depends on the
economic rationale for the existence and development of international firms, and the internal pricing policies adopted by the firms themselves.
4.b Economic Theory and Business Practice
Until relatively recently, economic theory did not concern itself specifically with the phenomenon of the transnational corporation. International economic relations were
treated as relations between national economies based on states, and consisting essentially of trade and investment and the corresponding international payments.
Trade theory assumed immobile production factors labour and capital, perfect competition, free access to technology, and no barriers to exports; it was
complemented by international investment theory which accepted the existence of trade barriers and the mobility of capital as a factor of production, explained as the
response to differences in price interest rates and resource endowments. Only from the late 1950s did international economists begin to try to explain the actual patterns
of international business, increasingly dominated by TNCs. They began to accept that `production factors
¶are much more diverse, and can vary significantly in quality e.g. labour skills; also, that markets are imperfect, and enterprises can protect
monopolistic advantages, especially in production technology and product specialization.
1
Hence, the existence of TNCs could be explained as the exploitation, on an
1
For a survey of the development of the debates see generally Dunning 1988 chapter 1.
international scale, of combinations of assets and locations which entail a competitive advantage. Initially, the focus was on firmspecific advantages of industrial
organization, protected by entry barriers, which gave foreign firms a monopolistic position compared with local firms Hymer 19601976, or technological advantages
deriving from their control over the diffusion of new products Vernon 1966. More recently the emphasis has been on analysing the advantages of the organizational
form, under the influence of institutional and transaction cost economics, which has tried to refine the analysis of the dynamics of the firm. Rejecting the basic
assumptions of the neoclassical paradigm based on the rational choice of the pure individual with perfect information, in favour of behaviourist and interactional
models which emphasise bounded rationality and the important social roles of informal and formal organization, institutional economics analyses the characteristics
of knowledge and its role in optimal decisionmaking, and argues that internalization of transactions within the firm may be more efficient in dealing with uncertainty, risk
and opportunism. Hence, the growth of the firm into the multidivisional corporation, the conglomerate and the TNC results not so much from economies of scale in
production which could be realised through contracting, but from the advantages of the organizational form, both in controlling and applying technological innovation
and in taking strategic capital investment decisions Williamson 1985, Ch.11.
Finally, writers have emphasised the advantages of spreading risks through international diversification Rugman 1979. In particular, TNCs can benefit from the
spreading of financial risks and the internalization of the management of arbitrage opportunities arising from diversity of currency exchange rates, financial markets and
regulatory regimes Lessard 1979. Important among these is of course taxation. This indicates that due to competitive pressure TNCs will seek the most favourable means
of structuring their internal financial and other transactions to take advantage of regulatory differences and the inadequacies of regulatory coordination.
Two main conclusions can be derived from these analyses. First, TNCs exist to a great extent because of the competitive advantages they can derive from internalizing
markets Rugman 1982. The economies derived from internalization may be considered to result either from reduction of production costs, whether through
economies of scale increased production volume of one item reducing its incremental or marginal cost or economies of scope reducing the marginal cost of
different items which can share joint factors; or from reducing transaction costs, by sharing and managing the costs and risks of appropriation of knowledge through
research, design and managerial and other professional services, and other advantages of coordinated organization. Both the production cost and transaction cost
perspectives emphasise that an integrated entity benefits from synergy profits, or additional returns attributable to oligopolistic advantages and the organizational form.
Second, the advantages they gain from internalization, which would not be
available to independent parties contracting through markets, entail the active management of internal transactions for the optimal benefit of the organization as a
whole. This is particularly the case for regulatory arbitrage.
It is this feature of strategic management of internal transactions that clearly emerges from the more specific theoretical and empirical studies of transfer pricing. Authors
have stressed that a companys internal pricing systems must respond to a variety of aims and constraints in relation to both its internal structure and goals as well as the
external environment Plasschaert 1980. It was not until the 1920s that large firms were forced to move away from centralized management structures to encourage
decentralization and allow vertical integration through divisionalization Chandler 1962, p.44. This entailed, at least for the large bureaucratic firms, notably DuPont
and General Motors, the `development of financial instruments which made it possible to establish decentralization with coordinated control
¶Sloan 1965, p.119. Such a financial system must reconcile the often conflicting requirements of
measuring performance and providing incentives for managers of costcentres. In order to establish criteria to judge return on capital invested for each cost centre, the
internal costing structure depends crucially, in modern largescale manufacturing, on assumptions about volume. Sloan emphasised that the key to the control system
developed by General Motors in the 1920s was the establishment of the standard volume concept, enabling the firm to ride out shortterm market fluctuations Sloan
1965, p.142148.
When, somewhat later, economists came to study the problem, they treated it as one in marginal analysis, and argued that unless perfect markets exist when market
prices should be used internal pricing should be based on marginal cost price Hirschleifer 1956, 1957. This has been cogently criticised from a business
management perspective, however, as focussing on profit maximization on the basis of highly restrictive assumptions. By assuming that the operating divisions have no
joint or common costs either for technology or marketing and that fixed costs are sunk, it omits the crucial element of business strategy, which entails choices and
motivation Eccles 1985. In practice, more recent surveys of business practice show that firms rarely use marginal cost, opportunity cost mathematically programmed or
even standard variable cost prices. Where a costbased price is used, it is full production cost, frequently with an allowance for profit also; furthermore, although
costbased prices are used less for international than domestic transfers, fullcost plus profit is more frequent internationally Tang 1981, esp. 141142. When asked to
rank the importance of factors affecting transfer pricing strategies, firms predictably place overall profit to the company first, while the competitive position and the
performance evaluation of foreign subsidiaries were very important; also highly placed factors were restrictions on repatriation of profits or dividends, the need to
maintain adequate cash flows, and devaluation and revaluation, as well as
differentials in income tax rates and income tax legislation between countries Tang 1981, p.144.
Hence, the internal transfer pricing policies of TNCs are defined by an evaluation of the various strategic factors affecting the firm, of which exploitation of opportunities
for financial, foreign exchange and tax arbitrage, within the constraints set by rule enforcement by state authorities, are among the most important external factors.
Thus, there are likely to be significant tensions between the policies of the firms in setting prices and the criteria to be used by the tax authorities in determining which
policies are acceptable. It is not clear that the independent enterprise standard is the best startingpoint for such regulatory criteria, since it cuts across the firms own
strategic perspective, and is therefore likely to exacerbate such tensions. As we have seen, it is where the authorities have moved towards acceptance of proportional
allocation of costs that there has been most harmony with company strategies. When it comes to the allocation of profits, although there has been apparent general
agreement in principle on the separate enterprise standard, its application in practice has tended to heighten conflict, both between firms and regulators, as well as between
national authorities among themselves.
Indeed, as the microeconomic methods applied to evaluation of transfer pricing by tax authorities have become more sophisticated, so they have increased these
conflicts. As has been shown above, the marginal analysis assumptions of conventional microeconomics ignore the central factors of oligopolistic markets and
organizational advantage; hence, their application will tend to result in the identification of a large item of residual profit Witte and Chipty 1990. The firms
own internal pricing strategy must take into account how to allocate this residual profit, balancing managerial incentives, performance evaluation, regulatory
compliance and avoidance and the other factors. However, neither managerial nor microeconomic theories can provide criteria for its interjurisdictional allocation.
The separate enterprise approach assumes in principle that, provided at least that foreign subsidiaries and branches show a reasonable return on assets or turnover, the
synergy profits should be attributable to the parent. Langbein has argued that a transaction costs approach, as opposed to one emphasising production costs, would
negate this assumption, since the returns to organization would be seen more clearly as returns to the organization as a whole Langbein 1989. However, transaction cost
analysis does not produce any better theoretical basis for interjurisdictional allocation, although it may be a better rationale for defining when a firm should be
treated as unitary see McLure 1984, p.103, and chapter 9 section 2 below. Indeed, Langbein concludes that the transactional analysis perspective, emphasising that
economies of integration result from management of risk, suggests an inter jurisdictional allocation based on a fractional system, `employing fractions designed
in accordance with a general understanding of the economic processes which produce and preserve the tax base
¶Langbein 1989 1410. However, beyond stating that these
fractions should be multilaterally agreed, he is not able to suggest how transactional analysis might produce appropriate allocation criteria. In principle, it is not clear that
the microeconomics of transactional analysis provides any better basis than that of production cost for proposing criteria for interjurisdictional allocation.
5. Conclusion.