Experimental Economics, Transport and Logistics
G de Jong, ITS, University of Leeds, UK and RAND Europe, NL
The paper will work out the parts of experimental economics theory that appear most relevant for application in freight transport. The design, conduct and outcomes of a logistics experiment will be given.
Experimental economics, transport and logistics
Gerard de Jong, Institute for Transport Studies, University of Leeds and RAND Europe
Experimental economics (EE) is a branch of economics that started in the late 1940s (but only became popular many years later) and centres around the use of laboratory experiments in which subjects, through some reward medium (usually cash) are induced to behave as economic agents. The main difference between EE and a stated preference survey is that in EE the reward mechanism is used to have respondents make choices according to the institutional rules of the experiments instead of according to his/her own ?innate? characteristics.
Types of applications; experience in transport
Davis and Holt (1993) list three types of applications of EE:
? Market experiments, in which the predictions of economic theory on market forms and price determination can be tested (e.g. perfect competition, oligopoly).
? Game experiments, in which the behavioural implications of specific games (prisoners? dilemma, sequential bargaining) are tested.
? Individual decision-making experiments, where the axioms of expected utility theory under uncertainty are investigated.
In transport analysis, individual choice theory has been applied extensively (though usually in the context of perfect knowledge of the decision-maker, not with uncertainty about the pay-offs). Game theory and market theory have also been used now and then, e.g. in the context of debates on deregulation (e.g. Preston, 1991) or network modelling. However, to the author?s knowledge, there are only a few applications of EE in transport. Isacsson and Nilsson (2004) performed experiments with students in Sweden on the allocation of rail track capacity to competing operators and Lunander and Nilsson (2004) on public procurement for road markings (renewing the paint) in Sweden. The only example in freight transport is Holguin-Veras and Thorson (2003), who had undergraduate students representing profit-maximising trucking companies.
Scope for application in freight transport
The use of EE in transport, and especially in analysing freight transport has considerable appeal. Two of the problems that freight modellers have to face are that for each shipment there can be several decision-makers (shippers, carriers, drivers, third and fourth-party logistics providers) and that data (especially at the level of the individual firm) are scarce. EE could help in solving both problems: it can be used to study situations with several interacting players (in a market and/or game-theoretic setting) and it can provide behaviourally rich data on commercial decisions for which data are usually not available. Especially market experiments could be particularly useful in freight transport, because many of the outcomes in freight transport are determined on markets (e.g. the freight rates that shippers pay for the services of the carriers).
Several market trading institutions have been investigated in EE. The most relevant candidates for freight transport analysis are the ?posted offer auction? and the ?double auction?. In the former type of auction, sellers propose prices publicly. Then the first buyer engages in as many transactions at the posted prices with sellers as he/she desires. After this, another buyer is selected at random and given a chance to buy. When all buyers have had this opportunity or when all sellers are sold out, the trading period stops. In a next trading period, the sellers can list new prices, etc.
In the double auction, there is more room for decision-making by the buyer, and the process of trading becomes sequential: in the experiment both buyers and seller can raise their hand and make public bids (buyers) and offers (sellers). In a trading period, the bids are raised and the offers lowered until the goods are sold (or a fixed amount of time has passed). This is the trading institution that is used most in EE, after the pioneering work by Vernon Smith (Nobel Prize winner in Economics in 2002 for his contributions in EE) in the 1960s, who found a fast convergence to competitive prices when using this mechanism for many situations.
A logistics experiment
At the University of Leeds a logistics experiment will be carried out, with students as the participants. In freight transport there are trade-offs between inventory costs and transport costs. Higher transport frequencies lead to higher transport costs (because of the use of smaller or not fully loaded vehicles), but smaller inventories at the retail end. This can be represented as a market game between receivers (retailers) on the one hand and senders (manufacturers) on the other hand. The receivers want to maximise the revenue from sales to consumers minus the sum of purchase and inventory costs. The senders want to maximise the revenues from selling their products to retailers minus sum of production and transport costs (and possibly inventory costs as well). The final outcome depends on the distribution of market power and access to information between the two sides, and several variants for these distributions will be tested (e.g. factory gate pricing, vendor-managed inventories, introduction of logistics service providers).
The paper will further work out the parts of experimental economics theory that appear most relevant for application in freight transport, focussing on market experiments to search for empirical regularities. The design and conduct of the logistics experiment will be described and the outcomes will be given in terms of institutional settings (cooperation and information exchange) that lead to efficient transport solutions (minimising vehicle kilometres) while avoiding excess profits.
Davis, D.D. and C.A. Holt (1993); Experimental economics; Princeton University Press, Princeton, New Jersey.
Holguin-Veras, J. and E. Thorson (2003); The role of experimental economics in freight transportation research: preliminary results of experimentation; paper presented at the 2003 European Transport Conference, Strasbourg.
Isacsson, G. and J.-E. Nilsson (2004); An experimental comparison of track allocation mechanisms in the railway industry; to appear in Journal of Transport Economics and Policy.
Lunander, A. and J.-E. Nilsson (2004); Taking the lab to the field: experimental tests of alternative mechanisms to procure multiple contracts; to appear in Journal of Regulatory Economics.
Preston, J. (1991); Explaining competitive practices in the bus industry: the British experience; Transportation Planning and Technology, 15, 277-294.
Association for European Transport