Reprinted by permission (with minor revisions) from Claude Ménard, editor (2000), Institutions, Contracts and Organizations: Perspectives from New Institutional Economics, Cheltenham, UK: Edward Elgar, pp. 367-375.

Alexandra Benham and Lee Benham

Introduction

Transaction costs fundamentally affect the functioning of an economy. They affect what is produced and what exchanges take place in the market; they affect which organizations survive and which rules of the game persist. Specific assumptions concerning transaction costs underlie the majority of theoretical models in economics, whether classical, Keynesian, neo-Keynesian, or neoclassical. Models dealing with monopoly, vertical integration, externalities, strategic behavior, wage and price stickiness, and imperfect markets all require specific assumptions concerning these costs. Transaction costs are arguably the most important set of prices in an economy. Yet how big are they? How much do they vary across settings and over time? Few empirical estimates exist concerning their magnitude and variation.

This chapter examines some conceptual and practical difficulties of estimating transaction costs empirically. We then consider a subset of the total costs incurred in a transaction, a subset which we designate the cost of exchange. The cost of exchange is defined as the opportunity cost faced by an individual using a given form of exchange to obtain a specified good within a given institutional setting. Some illustrative examples suggest an approach for estimating the costs of exchange empirically.

Elements of Transaction Costs

In modeling and in policy discussions, economists must deal constantly with the implications of different levels of transaction costs. Economists" theories of economic performance are based upon Adam Smith"s notion that wealth depends on specialization, which depends on the extent of trade, which depends on low costs of trade, that is, low transaction costs. Models in economics today include, explicitly or implicitly, specific assumptions about transaction costs. Neoclassical models assume they are zero. Neo-Keynesian models assume high transaction costs in some capital markets. Game-theoretic models assume high information and enforcement costs for some but not all contracts.

Imagine having ready access to credible information on transaction costs across time and across various institutional settings. Such information would be highly useful in providing answers to fundamental questions. What is produced? Who produces it? What exchanges take place in the market? Are there any poor economies with low transaction costs? Are there any rich economies with high transaction costs?

Currently there are few empirical estimates of transaction costs, and very few comparative estimates. The benefits of having better estimates are clear. Why does better information not currently exist?

First, there is no standard terminology. Many different definitions of transaction costs appear in the literature. They often serve as heuristic devices which are not used actually to measure transaction costs. These definitions offer powerful conceptual insights, but they have not been translated into widely accepted operational standards. Kenneth Arrow has defined transaction costs as "the costs of running the economic system." (1969, p. 48, as noted in Furubotn and Richter 1997, p. 40). Yoram Barzel defines transaction costs as "the costs associated with the transfer, capture, and protection of rights" (1997, p. 4). Thrainn Eggertsson observes, "In general terms, transaction costs are the costs that arise when individuals exchange ownership rights to economic assets and enforce their exclusive rights. A clear-cut definition of transaction costs does not exist, but neither are the costs of production in the neoclassical model well defined" (1990, pp. 14-15).

Eirik Furubotn and Rudolf Richter examine transaction costs in the following terms:

"[T]ransaction costs include the costs of resources utilized for the creation, maintenance, use, change, and so on of institutions and organizations….When considered in relation to existing property and contract rights, transaction costs consist of the costs of defining and measuring resources or claims, plus the costs of utilizing and enforcing the rights specified. Applied to the transfer of existing property rights and the establishment or transfer of contract rights between individuals (or legal entities), transaction costs include the costs of information, negotiation, and enforcement." (1997, p. 40)

"Typical examples of transaction costs are the costs of using the market [market transaction costs] and the costs of exercising the right to give orders within the firm [managerial transaction costs]….[There is also] the array of costs associated with the running and adjusting of the institutional framework of a polity [political transaction costs]….For each of these three types of transaction costs, it is possible to recognize two variants: (1) "fixed" transaction costs, that is, the specific investments made in setting up institutional arrangements; and (2) "variable" transaction costs, that is, costs that depend on the number or volume of transactions." (ibid., p. 43)

Second, estimation is problematic because production and transaction costs are jointly determined. This leads to formidable difficulties in estimating transaction costs separately. Economic theory suggests that changes in transaction costs have a first-order impact on the production frontier. Lower transaction costs mean more trade, greater specialization, changes in production costs, and increased output. Changes in production costs likewise have an impact on transaction costs.

Third, if transaction costs are very high, many kinds of transactions may not take place at all. Even when a specific kind of transaction does occur, it may not take place in an open market context with money prices. Hence of all potential transactions, only a small subset will actually occur, and only a subset of these will appear in the market. Ascertaining why a particular transaction is undertaken by an individual requires knowledge of the opportunity costs of alternatives. To understand the choices made, we may need to estimate the cost of transactions that did not actually occur.

Fourth, the law of one price does not apply here. Individuals and groups within a given society may face very different transaction costs, so many estimates may be needed. Other things equal, an individual"s political connections, ethnic group, and other characteristics will affect the opportunity costs of particular exchanges. These differentials are rarely transparent to outsiders.

Given all these difficulties, very few studies attempt to estimate these costs concretely. The private sector and organizations such as the World Bank and the United States Agency for International Development have been more active than the scholarly community in assembling data on transaction costs. This information has not been much used by scholars, partly because of confidentiality issues. Comparisons remain difficult because of the lack of agreed-on definitions.

Among the relatively few empirical studies is the work by Wallis and North (1986) examining historical trends in transaction costs in the United States. This study focuses on the overall size of the transaction sector and estimates that in 1970 about 45 percent of U.S. gross national product was devoted to the transaction sector. It does not estimate magnitudes for specific transaction costs. In a rare comparative study, Stone et al. (1996) compare business transaction costs in Brazil and Chile. Brazil"s legal and regulatory structure is more complex and less transparent than that of Chile, and the costs of the formal system appear to be higher in Brazil. However, in some arenas examined, Brazilian private institutions have evolved to mitigate these costs surprisingly well. Private facilitators provide a one-stop entry point for starting a new business; an innovative credit information system reduces the need to use the courts for debt collection. These results highlight the importance of observing actual practices.

The Costs of Exchange

We propose to examine a subset of the total costs incurred in a transaction, a subset which we designate as the cost of exchange. The cost of exchange C_ijkm is defined as the opportunity cost in total resources–money, time, and goods–for an individual with characteristics i to use a given form of exchange j to obtain a good k in institutional setting m. The cost of exchange is therefore the sum of production costs and the specific transaction costs faced by the individual. While we cannot decompose these costs directly into their production and transaction components, in comparative analysis this approach will focus attention on the total consequences of differing transaction costs.

This framework focuses on the opportunity cost faced by an individual seeking to enact a specified form of exchange (for example, via formal contract or informal arrangement, money or barter compensation) in a specified institutional setting. It does not include the costs of building market institutions, the costs of setting the political framework in place, or the cost to the individual of creating personal networks, establishing a reputation, or developing transaction-related skills.

To examine these costs empirically, we need a standardized methodology which specifies particular transactions in terms of the form of exchange, the good to be obtained, the characteristics of the individual, and the setting. Our approach is to select and specify some transactions in detail (for example, importing a crankshaft for a specified major brand of earthmoving equipment through the formal marketplace) so that researchers can measure the time and money costs incurred when the transaction takes place. Individuals with designated characteristics can then be interviewed concerning the time and money costs they have actually incurred in engaging in the transaction.

For many important exchanges, a conventional market price is not available, or if available is a very incomplete reflection of opportunity cost. In some cases, relevant information can be obtained through surveys of buyers, sellers, brokers, or facilitators who are directly involved in the exchange. In other cases, information must be obtained through simulations in which researchers go through the steps of an exchange themselves. This is particularly important if the exchange in question rarely takes place because the cost of exchange is so high.

Our initial choice of transactions has emphasized intermediate goods, such as obtaining a business telephone, transferring title to a piece of property, starting a new business, and importing a replacement part. Differences in the costs of exchange for these kinds of intermediate goods will substantially affect what is produced and how it is produced. To illustrate this perspective, we summarize below some examples taken from various countries in recent years. Some are drawn from our own investigations and some from work by others. These examples merely suggest the more systematic comparative work that will be possible in the future, given a more fully developed methodology and broader scope of survey.

Empirical Illustrations

Consider first the cost of obtaining a business telephone. This will affect the size of the telephone communications network, the extent of use, the overall size of the market, and the extent of specialization. In the early 1990"s, we investigated the cost of obtaining a business telephone in several countries. The actual price to obtain a telephone installed within two weeks ranged from $130 in Malaysia to $6000 in Argentina. In Egypt in l996, the official published price for a telephone was $295 and the official published "urgent response" price was $885. To proxy for the opportunity cost, we compared the purchase prices for similar Cairo apartments with and without a telephone already installed. This difference, which reflects the expected spot market price for a telephone for someone not well connected in this market, was approximately $1180 to $1770.

Efficiently transferring ownership of assets is fundamental to a modern market economy. The costs associated with transferring ownership of an apartment can be examined in this context. In Cairo, an individual who buys an apartment and registers the transfer of ownership pays an additional 12 percent of the apartment price to third parties: 6 percent for taxes and 6 percent for a lawyer to register the transfer as required by law. The services of a real estate agent, which are optional, cost about 1.5 percent of the sale price. In St. Louis, Missouri, USA, the cost of legally transferring ownership is approximately 1.5 percent of the sale price; if the services of a real estate agent are used there, they cost 6 percent of the sale price. The differences across these rates are striking. Fees are eight times as high in Cairo as in St. Louis within the state-controlled sector, but only one-quarter as high within the competitive sector (Benham 1997).

Obtaining legal permission to open a new business is another arena of interest. An example of the simulation approach is provided in Hernando de Soto"s book, The Other Path (1989). In Lima, Peru, in 1983 a team of researchers undertook the bureaucratic procedures necessary to establish a new small garment factory legally. They attempted to carry out all procedures without paying bribes (only twice were bribes unavoidable) or using political connections. Detailed notes and time logs were kept. The results showed that a person of modest means would have to spend 289 days in these procedures to set up the factory legally. It is not clear that anyone in Peru other than de Soto"s team ever went through this entire process to obtain a permit, but this estimate of opportunity cost was entirely consistent with the choices individuals were making at that time. Those without political connections typically remained in the informal sector, not legally registered. When de Soto repeated the simulation in Tampa, Florida, it took only two hours to obtain a permit to open a small business. Thus in Peru the time cost was over 1000 times as high as in Florida. In this kind of highly bureaucratized environment, the costs of not being physically located in the capital city can be daunting. For example, in Tanzania a business partnership based in Mwanza outside the capital must spend five to ten times as much to register as a business partnership based in Dar-es-Salaam (The Services Corporation 1998).

Transactions across national borders are important indicators of the extent of the market. To look at variation across countries, we examined the cost of exchange associated with importing a crankshaft for a large earthmoving tractor. In Peru in 1989, formally obtaining this crankshaft cost four times as much in money price and over 280 times as much in waiting time (41 weeks versus one day) as in the USA. In Argentina, the money price was twice that in the USA, and the waiting time was up to 30 days. In contrast, in Malaysia the money price and waiting time were essentially the same as in the USA. In Hungary, before currency and import regulations were liberalized around 1989, it took 30-48 weeks to replace a crankshaft for a western-made tractor; after liberalization, this wait dropped to two weeks. A related measure is the average waiting time to clear items already in port. In Singapore this is 15 minutes, while in Tanzania it is 7-14 days, with waits of up to 91 days reported (ibid.). Fourteen days" wait is more than 1300 times the average waiting time in Singapore.

What do these examples tell us about the costs of exchange? First, the variation in the costs of exchange is enormous. We observe variations as large as one to three orders of magnitude. Such great variation is seldom seen in the published data on money prices. For example, in developed countries, monopolies rarely sustain long-run prices more than 20 percent above the competitive level. Second, the full opportunity cost which individuals incur in obtaining a good may differ greatly from the conventionally observed money price. The published price data which economists typically use in empirical work may poorly represent the opportunity costs facing individuals who are deciding what exchanges to undertake.

Building Future Research

The opportunity cost of measuring the opportunity costs of exchange is likely to be high. This is particularly true when simulations are undertaken. As noted, de Soto"s research team spent over 289 days simulating the process of registering a new business in Peru. Published data seldom provide the information needed. Sometimes clever proxies can be found for these costs, but not always. Nevertheless, we believe the payoff will be great if there are more systematic collaborative efforts across countries using this approach. The investigations already conducted indicate that such an approach is feasible. The very substantial variations shown above in the costs of exchange suggest that it is important.

Not all research questions can be addressed by one strategy, but the recommendation here is to develop a framework for collaborative analysis which involves developing common operational definitions and a common research protocol. Research groups in different countries could then focus on estimating the costs of exchange sector by sector in their own countries. These estimates could be combined for comparative analysis across groups, countries, and time periods. If the price of an intermediate good needed for final product is ten times higher in country A than in country B, we should not be surprised if country A does not produce that final product.

With the accumulation of such information, one could map out and compare the extent of the market across settings with different costs of exchange. How far can an individual with given characteristics trade across space, over time, and across other parties? Where contract enforcement is very expensive, exchange will be concentrated within sets of individuals who know and trust each other. Where it is less costly, the exchange frontier can expand from the highly personal to the impersonal, to include extended family members, clans, social groups, national groups, even foreigners. Similarly, where information, measurement, and enforcement costs are lower, exchange can take place over greater distances and longer time periods. These variations have implications for the degree of specialization and ultimately the performance of the economy.

Bibliography

Alston, Lee J., Thrainn Eggertsson, and Douglass C. North (eds) (1996), Empirical Studies in Institutional Change, Cambridge: Cambridge University Press.

Arrow, K. J. (1969), "The Organization of Economic Activity: Issues Pertinent to the Choice of Market versus Non-Market Allocation," in The Analysis and Evaluation of Public Expenditures: The PBB-System, Joint Economic Committee, 91st Congress, 1st session, vol. 1, Washington DC: Government Printing Office.

Barzel, Yoram (1997), Economic Analysis of Property Rights, 2^nd edn, Cambridge: Cambridge University Press.

Benham, Alexandra and Lee Benham (1989), "Survey on Costs of Exchange," unpublished manuscript.

Benham, Alexandra, Lee Benham, and Michael Merithew (1995), Institutional Reform in Central and Eastern Europe: Altering Paths with Incentives and Information, San Francisco: International Center for Economic Growth.

Benham, Alexandra and Lee Benham (1997), "Property Rights in Transition Economies: A Commentary on What Economists Know," in Nelson et al. (1997), pp. 35-60.

Benham, Lee (1997), On Improving Egypt"s Economic Performance: The Costs of Exchange, Cairo: Egyptian Center for Economic Studies, Working Paper No. 13.

Coase, Ronald H. (1937), "The Nature of the Firm," Economica 4, 386-405.

Coase, Ronald H. (1998), "The New Institutional Economics," American Economic Review 88(2), 72-74.

de Soto, Hernando (1989), The Other Path: The Invisible Revolution in the Third World, New York : Harper & Row.

Eggertsson, Thrainn (1990), Economic Behavior and Institutions, Cambridge: Cambridge University Press.

Engerman, Stanley L. and Robert E. Gallman (eds) (1986), Long-Term Factors in American Economic Growth, Studies in Income and Wealth, No. 51, Chicago and London: University of Chicago Press.

Furubotn, Eirik G. and Rudolf Richter (1997), Institutions and Economic Theory: The Contribution of the New Institutional Economics, Ann Arbor: The University of Michigan Press.

Nelson, Joan M., Charles Tilly, and Lee Walker (eds) (1997), Transforming Post-Communist Political Economies, Washington, D. C.: National Academy Press.

North, Douglass C. (1990), Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press.

Shelanski, Howard and Peter Klein (1995), "Empirical Research in Transaction Cost Economics: A Review of the Evidence," The Journal of Law, Economics and Organization, 7(2), 335-361.

Stone, Andrew, Brian Levy, and Ricardo Parades (1996), "Public Institutions and Private Transactions: A Comparative Analysis of the Legal and Regulatory Environment for Business Transactions in Brazil and Chile, " in Alston et al. (1996), pp. 95-128.

The Services Corporation (1998), USAID Executive Summary, The Investor Roadmap to Tanzania, Report.

Wallis, John J., and Douglass C. North (1986), "Measuring the Transaction Sector in the American Economy, 1870-1970," in Engerman and Gallman (1986), pp. 95-161.