Graduate optional module
10 credits, no pre-requisites
Course Organiser: John D. Hey
(One term module – Autumn Term 2000)
PREAMBLE
This module should involve you in about 100 hours work, some individual, some in groups. The aims of this module are to introduce you to the methodology, philosophy and practice of Experimental Economics. Its learning objectives are to get you to be able to criticise constructively the experimental work of others and to carry out your own experimental work yourself, whilst at the same time getting some knowledge of what experimental work has been done by others and the contribution of that work to the science of economics. The structure of the module is described below; in summary it consists of 9 lectures describing, discussing and evaluating experimental work in 9 different areas, all involving some kind of interactive economic behaviour. The teaching methods used in the course involve a combination of lectures and interactive seminars. The assessment is by examination and is described in more detail below. The reading required is also detailed below.
INTRODUCTION
The course this year is completely re-designed. I am taking advantage of the new EXEC laboratory (equipped with 15 networked Windows NT PCs) and the availability of the experimental software Z-Tree (written by Urs Fischbacher at the University of Zurich) to offer a course with a substantially increased ‘hands-on’ element – so that students can get experience, not only of participating in experiments, but also of writing and running their own interactive experiments using Z-Tree. This latter will involve a modest amount of relatively simple ‘program writing’ as well as serious involvement in designing, planning, testing, implementing and analysing experiments. It should be fun.
Coincidentally with this new type of course, I am also changing the actual topics studied. Compared with previous years, there is now much more emphasis on interactive models of economic behaviour. We will study theoretical solutions to the various interactive problems and examine experimentally how well these theoretical solutions perform in the laboratory. The course consists of 9 lectures and 4 seminars. The lectures are weekly, in Weeks 2 through 10, and the seminars fortnightly, in Weeks 4, 6, 8 and 10. The lecture titles are as follows:
The seminars will all be practically based. We will meet in the EXEC laboratory (Derwent D block room 203). You will need the D-block code to enter the staircase – you should ask the Derwent porter. The seminar titles are as follows:
Whilst I will give a basic introduction to Z-tree, and give you some examples, you will be expected to devote a considerable amount of time to familiarising yourself with the software. You could do this individually, but I would encourage you to work in groups. For the purpose of the seminar work, you will be organised in groups and you are encouraged to use these groups in an efficient fashion – appointing experts to operate on certain aspects of the group work. You will not be expected to write Z-Tree programs in the examination. The Z-Tree manual is downloadable from Urs Fischbacher’s Z-Tree web site – at http://www.iew.unizh.ch/ztree/. If you visit this site you will also see that Urs runs a mailing list on which you can enter into interesting discussions with other experimental economists. Apropos of that, you might like also to register on the Experimental Economics Mailbase list (which I run) at http://www.mailbase.ac.uk/lists/economics-experimental/. You might also find of interest the web site of the Economic Science Association (the rather grandly titled association of experimental economists throughout the world) – of which I am proudly the European Secretary at the moment – at http://www.econ.nyu.edu/Dept/ESA/index.htm/.
GENERAL READING
As far as reading is concerned, it might be helpful to first discuss general sources of bibliographical information about experimental economics. I list this under two broad headings – first textbooks, second overviews, surveys and collections of readings. As far as the first heading is concerned, there are now three text books on experimental economics available (though the third is out of print in its English version - but there are Spanish and Italian editions available):
Davis, D. D. and Holt, C. A. 1993), Experimental Economics, Princeton U.P.
Friedman, D. and Sunder, S. (1994), Experimental Methods: A Primer for Economists, Cambridge U.P.
Hey, J. D. (1991), Experiments in Economics, Blackwell.
They each have their merits (and demerits), though the emphasis varies from text to text (for example: Hey is more concerned with individual experiments; Davis and Holt with market experiments; and Friedman and Sunder with (practical) methodological issues). You may find it useful to buy one or other of these, though this is not essential: copies will be available in the library.
The second heading includes the following sources:
Hey, J. D. and Loomes, G. C.(1993), Recent Developments in Experimental Economics, two volumes, Edward Elgar.
Kagel, J. H. and Roth, A. E. (1995), Handbook of Experimental Economics, Princeton U.P (now in paperback).
Roth, A. E. (ed.) (1987), Laboratory Experimentation in Economics: Six Points of View, Cambridge U.P.
Roth, A. E. (1988), "Laboratory Experimentation in Economics", Economic Journal, 98, 974-1031.
Smith, V. L. (1990), Experimental Economics, Edward Elgar.
The Kagel and Roth Handbook is particularly useful for anyone wanting an overview and a survey of the key areas within economics. It contains a comprehensive bibliography and provides a very informative and stimulating introduction to the relevant literature. It is invaluable if you are planning to do research involving experimental work yourself. There is now a new journal, Experimental Economics, devoted to work in this field.
You are invited to read as much or as little of the above references. What is crucial is that you read at least the references listed below. I have tried to make the reading as straightforward as possible – though you are encouraged to read more widely. The references contained in the Handbook of Experimental Economics (Kagel and Roth (1995)) need not be read in detail – these are mainly to give you the appropriate background and a source of reference if you wish to take your study further. But the references marked with an asterisk* must be read. I would also advise you very strongly to read the Introduction by Roth in Kagel and Roth (1995).
THE SPECIFIC LECTURE READING
Roth A.E., "Introduction", pages 26 to 28 only for this topic, in Kagel and Roth (1995).
*Selten, R. and Stoecker, R. (1986), "End Behavior in Sequences of Finite Prisoner’s Dilemma Supergames: A Learning Theory Approach", Journal of Economic Behavior and Organization, 7, 47-70. (reprinted in Hey and Loomes (1993)).
Ochs, J. (1995), "Coordination Problems" in Kagel and Roth (1995).
*Cooper, R., DeJong, D., Forsyth, R. and Ross, T.W. (1990), "Selection Criteria in Coordination Games: Some Experimental Results", American Economic Review, 80, 218-33. (Reprinted in Hey and Loomes (1993)).
*McKelvey, R.D. and Palfrey, T.R. (1992), "An Experimental Study of the Centipede Game", Econometrica, 60, 803-836.
Ledyard, J.O. (1995), "Public Goods: A Survey of Experimental Research", in Kagel and Roth (1995).
*Andreoni, J. (1988), "Why Free Ride?: Strategies and Learning in Public Goods Experiments", Journal of Public Economics, 37, 291-304. (Reprinted in Hey and Loomes (1993)).
Burlando, R. M. and Hey, J. D. (1997), "Do Anglo-Saxons Free-Ride More?", Journal of Public Economics, 64, 41-60.
Nagel, R., (1995), "Unravelling in Guessing Games: An Experimental Study", American Economic Review, 85, 1313-1326.
*Ho, T. H., Camerer, C. and Weigelt, K., (1998), "Iterated dominance and iterated best response in experimental "p-beauty contests"", American Economic Review, 88, 947-969.
Banerjee, A. (1992), "A Simple Model of Herd Behavior", Quarterly Journal of Economics, 107, 797-817.
*Allsopp, L. and Hey, J.D. (200), "Two Experiments to Test a Model of Herd Behaviour", Experimental Economics, forthcoming. (Available as University of York, Department of Economics, Discussion Paper 99/24, and hence downloadable from the University website.)
Anderson, L. R. and Holt, C.A. (1997), "Informational Cascades in the Laboratory", American Economic Review, 87, 847-862.
Holt, C.A. (1995), "Industrial Organization: A Survey of Laboratory Research", in Kagel and Roth (1995). (Only part of this is directly relevant to the lecture topic, but all of it is interesting.)
*Smith, V.L. (1962), "An Experimental Study of Competitive Market Behavior", Journal of Political Economy, 70, 111-137. (Reprinted in Smith (1990)).
Hey, J.D. and Di Cagno, D. (1998), "Sequential Markets: An Experimental Investigation of Clower’s Dual Decision Hypothesis", Experimental Economics, 1, 63-86.
Sunder, S. (1995), "Experimental Asset Markets", in Kagel and Roth (1995).
*Smith, V., Suchanek, G.L. and Williams, A. (1988), "Bubbles, Crashes and Endogenous Expectations in Experimental Spot Asset Markets", Econometrica, 56, 1119-1151. (Reprinted in Hey and Loomes (1995)).
*Plott, C. R. (2000), "Markets as Information Gathering Tools", Southern Economic Journal, 67, 1-15.
THE SEMINAR PROGRAMME
This year the seminar programme will involve a much more active input from you than has been the case in previous years. Part of the seminar programme will be to show to you other people’s experiments and the other part will be for you to learn how to design and implement an experiment of your own. To this latter end you will divide yourself into groups (the precise number and size depending on how many students are following the course) and the ultimate aim of each group will be to design, construct and implement (on the other members of the course) your experiment. Ideally, groups will choose their own experimental topic, but if you are unable to come up with a topic I will suggest possible topics. The actual seminar programme will be as follows. Note that all seminars will take place in the EXEC laboratory.
In this you will largely be passive. I will outline Z-tree and guide you through some simple examples. You should come prepared having read as much as possible of the Z-tree manual. I will make available one copy for each group. You should try to form yourself into groups before this seminar.
In this seminar you will be shown a number of Z-tree experiments written by other experimentalists. You will play at being subjects. The seminar will conclude with a discussion of the design of the experiment and the construction of the Z-tree programme.
In this seminar, each group will outline its plans for its experiment, first describing what it is that they wish to investigate and how they intend to implement the experiment. There will then be a general discussion of these plans so that each group can go away from the seminar with a clear idea of how they will actually carry out the implementation of the experiment. Groups should come prepared, producing in advance a one-page outline of their experiment which can be distributed to the other participants in the course.
The period between the third seminar and the fourth will be an intensively active time for you. In this period you will need to write an appropriate Z-tree programme, draft some instructions for your subjects and give some thought to the choice of any relevant parameters. At this fourth seminar, each group will implement their experiment using the other members of the course as subjects (though no payments will be made). You will need to give me your Z-tree program some 24 hours in advance of the seminar so that I can load it on the EXEC net.
THE EXAMINATION
Because of the changed structure of the course this year, the examination this year will be different from that in previous years. It will be extremely straightforward: you will be asked to (individually) comment on the experiment that your group has designed and partially implemented. In particular you will be asked to motivate it, describe it and discuss its strengths and its weaknesses. The precise wording of the examination question can not be given at this stage, as it needs to be approved by the External Examiner, but nevertheless you can prepare for it in general terms. You will, of course, be expected to know how your experiment relates to, and improves on, the existing literature, and be prepared to criticise it in a constructive fashion.
A specimen examination paper follows. You will have two hours to answer the question.
"Describe the experiment that your group worked on during the course. Discuss its relationship to the existing literature and describe how you think your experiment may contribute to knowledge in that field. Pay particular attention to the difficulties involved in implementing the experiment and discuss ways that you think your design has overcome these difficulties. What further improvements and modifications do you think might be useful?"
OTHER
I hope you enjoy the course. Your comments on it will be welcome. You are strongly advised to be active throughout the course as much will depend on the interaction between me and the students on the course. I have a part-time position in York and teach part of the year in Italy. I leave the UK at the end of February and do not return until June, though I can be contacted via e-mail on jdh1@york.ac.uk throughout the year.
John Hey
23 October 2000