Posted by Jan-Eric Nilsson on December 12, 1997 at 08:54:55:
Jan-Eric Nilsson, Erika Budh
CTS 1997-12-12
THE USE OF ECONOMIC EXPERIMENTS IN BASIC LEVEL TEACHING AT DALARNA UNIVERSITY
During September and October 1997 (9 weeks) we have been using experiments in the first-year course in micro-economics at Dalarna University. This mimeo reviews the organisation of the course and observations related to the use of experiments.
THE GROUP
Approximately 100 students registered and about 80 of them completed the course. For most of them, 20 points in economics is compulsory to take out a BA exam. Perhaps 60% of the group follow a ‘teaching programme’ which emphasises business administration. (Most of them take a course in business administration/marketing parallel to the one in economics.)
TEACHING, LITERATURE
About 30 hours of standard lectures were given, primarily based on Parkin Economics (third edition), chapters 1, 2, 3 (not pp 53-58), 4-6, 9, 10 (not appendix), 11, 12, 13 (not pp 303-304, pp 311-321), 14 (not pp 345-347), 15 (only pp 358-360), 18, 20 (not pp 480-483) and ch 21. This means that we have jumped significant parts of a standard first-year course, in particular the demand-side analysis. Other sections were given only superficial coverage.
Another 6 hours/three ‘exercise meetings’ were used to help students to solve a number of exercises. These are fairly closely related to the questions in the final exam. These tasks are included in a compendium which also includes a number of self-study tasks.
6 experiments from B&M were also used, viz. Ch 1, 2, 3, 8, 7, 6 in the order now given. Most of the time the experiment was run before the lecture, once a week during the first four course weeks and week number six and seven. All experiments were run on Wednesdays and (a translated and rearranged version of) home work was handed out immediately afterward. Students participated in one out of three groups, i.e. each group comprised between 20 and 30 students. Homework was to be done in groups of 1-3 students and handed in for grading no later than Monday noon. Graded home work was returned to students on Tuesday’s lecture.
There was a morning lesson from 8-10 AM, another from 10-12 AM and an afternoon lesson from 1-3 PM. All experiments, including (a short) discussion plus distribution of home work material etc., were completed in less than one and a half hour, several even in about one hour.
All experiments had the following in common: Students were recommended to read through the experiment before coming to class (less than 50% of the group did so!). The experimental sessions commenced with a brief verbal repetition of rules and recommendations. Yellow slips of paper with ID number were handed out to simplify registration of trade and profits. Students wrote down their names and the Swedish correspondence to Social Security Number. The slips were collected after the experiment.
The copies of types from the book were cut into the number of slips that there were sessions in the experiment. The slips were packed for the different sessions following the recommendations in the teacher guide.
We started the first period of the first session with handing out private information. All slips with private information were collected after the first period, shuffled and than returned to the students for the second period of the first session. In the second session we used a the set of type packets for that session in the same manner.
Rules for trade: No communication between students were allowed. Students remained seated. Those who liked to place or accept a bid or ask raised a hand and waited to be called upon to speak by the experimenter. When called upon, (s)he placed a bid or accepted a bid; f ex ‘2 sell for 25’, ’12 buy for 18’, ’14 accept to sell for 18’. All bids, asks and trades were registered on the white board in the way shown below, to make the information available to everyone simultaneously. The number within parenthesis is the ID number.
BUY SELL
(2) 25
(12) 18
(1)accept
When a buyer and a seller reached an agreement, they left their slips with personal information to the assistant who registered the trade on a lap-top in Excel.
The first two experiments were compulsory; anyone not participating in these will have to complete a major home-work before they pass the course. For anyone who did not participate in any of the remaining four experiments, the requirement for Pass/Very Good on the exam - i.e. for final grade on the course - increases with 0,5 p. The total number of points in the exam were 30, 15 p were required to pass and 22,5 for Very Good. Anyone who had not participated in neither of the four exercises would have these boundaries changed to 17/24,5. Moreover, we kept a record of the students’ results so that it could be established who had succeeded best over all six experiments. #1 received a remuneration of SEK 250 , #2 SEK 150 and so on; in total eight prices were awarded.
PARTICULAR COMMENTS
Demand and Supply, Shifting Supply - Our first two experiments turned out (surprisingly) well.
A Sales Tax - Turned out well in group number 2 and 3. When starting the first session in the first group we had 23 students present, during the session number 24 showed up. She was allowed to join from session number two, which turned out to be a fatal mistake. Changing the number of participants from 23 to 24 changed the structure of the experiment in a way that destroyed the whole point of the experiment!! The theoretical sum of consumer and producer surplus in the second session exceeded the same sum for the first session. There were as well two late arrivals in the afternoon group as well, but this did not affect the experiment structure (number of participants 26+2). We won’t let late arrivals join our experiments again; but breaking the ‘package structure’ in an experiment seems to be crucial. This experiment was based on 12-packs, adding one to 23 meant using two full packets instead of one and adding two to 26 did not affect the number of packets used. Though the experiment result did serve as an excellent demonstration of the ceteris paribus assumption.
Entry and Exit- - This experiment got closest to failure due to sensitive design. One student that didn’t understand the incentives (and continued to open restaurant regardless of number of restaurant all ready in the market) were able to impose negative profits on those restaurant owners that did understand. In all three groups this experiment required continuous instructions and ‘leading’ by the experimenter. The morning experiment turned out least well, probably because it took a while to find out how to guide students ‘toward good decisions’. Good in the sense making the experiment serve it’s purpose.
Monopolies and Cartels - We made a few design changes, enclosed on the following page. Three sessions were run, the second and third session were probably the most entertaining ones during the whole course. Sellers were determined by lottery and taken out side to decide on strategy, with guidance of one of us. The other one of us stayed with the buyers and discussed possible strategies with them. The sellers sat next to each other and acted together during the period for which they were sellers. No one was allowed to be a seller twice.
Externalities - In the third session (with tradable pollution permits) we recorded the two markets parallel on the board, traders told which market (s)he wished to act upon and then placed or accepted a bid. Also this experiment work out well in all groups.
GENERAL EXPERIENCES
The experiments were a great success, an impression confirmed by the students. In particular, experiments in teaching provide them with a hands-on experience related to the way that the market works. We believe that this make (some of) them realise that Economics actually is more than abstract theory. I have been teaching about six groups over the years on first-year courses, and I strongly believe that this group more than others will understand where the X in economics (crossing Supply and Demand) comes from, and also remember it’s implications for the future.
The major complaint that we had from the students was that in particular the first two hand-ins took much too much time to complete. This may have been a correct observation, and we intend to try to be a bit more precise in the way that questions are formulated, possibly also to give them some extra hints about what to do and possibly also reduce the number to tasks slightly next year. Primarily, however, the problems are probably explained by a threshold which they must climb over before they fully understand what training the experiments give them.
We have been using open outcry trading, with bids, asks and deals continuously noted on the board. During a trial last year we applied the circulation approach suggested in the book, but our experience is that the dispersion of information works better when everything is publicly announced. In particular, prices converge towards equilibrium predictions much faster.
Since this was the first time the technique was used in a ‘full-scale application’, we were able to hire an extra assistant for the experiments. This meant that we were three personnel during the experiments; Nilsson working as the market maker, Budh taking notes on the board and the assistant registering deals and redemption values/production costs on a laptop. After one experiment we realised that this made it feasible for students to download the trade on a diskette; this opportunity was used by about 25% of the group. Next year we will strongly encourage them to use this possibility and to use standard software (Word and Excel) for hand-ins.
Our opinion is that one of the major advantages with using experimental techniques is that it provides students with intuition about the significance of our subject. At least in Sweden, we have a great number of students who voluntarily or compulsory take a 20 point course in economics and then nothing more. It is very important that these are provided with a proper understanding about how the market actually works, as discussed by the editorial in the Economist in September.
But the use of experiments is also valuable to students who continues to study economics. One concern that one can have over micro-courses thought to non-graduates is that they are extremely similar, differing only with respect to the extent of mathematical sophistication used. A first term micro course emphasising intuition is conveniently complemented by second and third terms which can be both deeper and wider in scope; this would further develop the way that we teach students.
A couple of comments that we picked up after that the course was completed made us realise that we have provided sufficient scope for discussion of results. Next year, lecture hours will be slightly increase in order to care also for this. Our preliminary intention is also to include another two experiments, viz. #4 and 5; the reason is that this will further add to their intuition.