The Journal of African Economies has just published its November issue, which is a special issue on impact evaluation.
Here is the table of contents. Here is the introductory essay by Marcel Fafchamps and Andrew Zeitlin, who write:
Two features are evident from the collection of papers presented here. First, as illustrated by the diversity of topics covered in this volume, evaluation methods can be applied to a broad range of policy questions. Such questions range from microeconomic and localized policies, such as in health and education, to policies with potential for general equilibrium and market-wide effects, such as migration and entrepreneurship.
I was in Montreal two weeks ago to take part in the 2012 McGill Global Food Security Conference, presenting in a session titled “Can High Food Prices and Volatility Be Managed?” Here are my slides (document opens a .pdf file).
My answer to the question posed by the session’s theme was: “Yes, but high food prices and food price volatility are different problems requiring different policy instruments,” and most of my presentation was dedicated to explaining (and then showing empirical examples of) how “food price volatility” is not the same thing as “high food prices.”
Given the popularity of my post on the trading game earlier this semester, I thought I should discuss another experiment I run in my principles of microeconomics class to get them to learn the rudiments of economics. This one is even simpler to run–all you need is a deck of cards.
Once again, I didn’t invent anything, as I rely on Charles Holt’s instructions for a market experiment, which you can find here (pages 2 to 5; link opens a .pdf document).
The idea is pretty simple: I split the class into two groups (i.e., buyers and sellers), and I assign each buyer a willingness to pay (WTP) and each seller a willingness to accept money (WTA) for one unit of some imaginary commodity. Those values must be kept secret.
Once students have been assigned to buyers or sellers and have received their valuation for the imaginary commodity, I tell them that they have five minutes to go out there and try to find the best deal for themselves.
For example, if a student is a buyer whose WTP is $5, she should only accept trading offers from sellers whose WTA $5 or less. Indeed, if she meets a seller whose WTA is $6, the minimum acceptable price for that seller exceeds the buyer’s maximum acceptable price, and no trade occurs.
Ideally, a buyer should seek to pay the lowest possible price, and a seller should seek to receive the highest possible price.
For example, if our buyer whose WTP is $5 meets a seller whose WTA is $2, the buyer wants to pay a price that is as low as possible, and the seller wants to receive a price that is as high as possible. Given their individual valuations, however, the only range of price for which a trade is possible is the $2 to $5 range.
Valuations have to be kept secret all along so as to make sure that trading parties do not take advantage of one another. This is like when you buy a house: You don’t start with your highest offer, and you make sure not to reveal your true valuation to the seller to make sure that you pay as little as possible for the house.
What’s in a Price?
Here is the really important part of the experiment, however: As trades occur, the price at which each trade occurs is recorded by the experimenter and announced to the whole class. This is so remaining buyers and sellers know what price is feasible. In other words, to simulate the fact that on most markets, prices are known and serve as a useful signal to market participants.
When I run that experiment, we usually play three to five rounds of the game. For each new round, I assign students to a group different from the one they were previously assigned to, and I give them a new valuation.
If you use a specific valuation structure, it is relatively easy to plot what supply and demand would look like. I use the exact valuation structure found in Holt’s instructions, so my experimental market should look like this:
(Source: Holt and McDaniel, 1996).
And lo, we usually hit the $5- to $7-price range as early as the first round!
Once we are done trading, I show students the graph above to show them that one could predict their (aggregate) behavior. This is usually when students truly get how markets operate, and how prices are determined by the interplay between buyers’ WTP and sellers WTA for a commodity.
Next year, I am thinking of tweaking the experiment by making it incentive compatible by giving students one unit of something relatively cheap (e.g., candy) for each unit of surplus they get. Thus, if our buyer whose WTP is $5 meets a seller whose WTA is $2 and they agree on a price of $3, the buyer would get two pieces of candy (WTP – Price = $5 – $3 = $2), and the seller would get one piece of candy (Price – WTA = $3 – $2 = $1).
