Economics Games

Classroom Games For Teaching Economics

New : The Bubble Classroom Game


This Thursday, new classroom game about speculative bubbles, by Sophie Moinas and Sébastien Pouget on, in the finance section.



Abstract of the paper:

“Students sequentially trade an asset which is publicly known to have a fundamental value of zero. If there is no cap on asset prices, speculative bubbles can arise at the Nash equilibrium because no trader is ever sure to be last in the market sequence. Otherwise, the Nash equilibrium involves no trade. Bubbles usually occur with or without a cap on prices. Traders who are less likely to be last and have less steps of reasoning to perform to reach equilibrium are in general more likely to speculate.”

(SEJ 2016:

Next Thursday, we will add a simulation introducing the perfect competition model.

facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

New: The Herd Immunity Game


This week, we are adding a new free game to the site: the Herd Immunity Game, by Alan Grant, Jim Bruehler and Andreea Chiritescu (Journal of Economics Teaching 2016). You will find it in the section “externalities and public goods” on this page: .



“Outbreaks of dangerous, preventable diseases have drawn attention to individuals who fail to obtain available and effective vaccines. This classroom experiment demonstrates the basic cost-benefit tradeoff inherent in vaccination. As more students obtain a costly vaccine, the likelihood of a non-immunized student catching the disease declines; non-vaccinating students obtain herd immunity. In equilibrium, a substantial fraction of students fail to obtain the vaccine. In addition to highlighting a genuine public health issue, the experiment can also be used more generally to illustrate the nature of externalities and the public goods problem.”


The paper is available on the site of the (open) journal:

Next thursday, we will add “The Bubble Game”, by Sophie Moinas and Sébastien Pouget (Southern Economic Journal 2016,, a classroom experiment about speculative bubbles.

facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

New oTree games (Auctions, Principal-Agent, Market for lemons…)

We are happy to announce that we have integrated oTree inside our site,

oTree is a great free open source software for creating economics experiments and classroom games, which is already used in many experimental economics research centers (, “oTree – An open-source platform for laboratory, online, and field experiments”, Journal of Behavioral and Experimental Finance, vol. 9, n° 1, 2016, p. 88–97, by Daniel Chen, Martin Schonger, and Chris Wickens).


matching pennies

If you have a little time, we would highly recommend you to code your own game (in Python): Everything is very clearly explained on the oTree website.


If you do not want to host your own server, you will find here several of the introduction games that are delivered with oTree:

  • Market for Lemons
  • Vickrey Auctions
  • Common-Value Auctions
  • Principal-Agent game
  • Bertrand Competition
  • Beauty Contest
  • An Ultimatum game
  • A Matching Pennies game
  • The Traveler’s Dilemma
  • A Trust game
  • A Stag/Hunt game
  • A Bargaining game

More advanced games, including original ones, will be added on a regular basis.

Also, If you have coded your own game or experiment and feel like publishing it in open access on our platform (or if you want to share a pedagogical document about existing games), do not hesitate to contact us.


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

Intermittent production, correlated shocks, constant feed-in premiums and price volatility

Taken from the last session of the Energy Game.

Here is what happens when off-peak demand is even lower than expected (-10%), while wind blows more than expected around wind farms (+27%). As a constant feed-in premium made it interesting to produce power from wind farms even at a null price, price on the wholesale market dropped to a very low level (2€ / kWh).

pricenullwindThis is even more dramatic later in the game, when flexibility features of power plants are taken into account.

In the next version of the game, it will be possible to have negative prices on the wholesale markets.


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

Stackelberg and Vertical Differentiation

In this example taken from our last game session at ENAC, one team (Flight club) used the option to precommit and chose to operate flights with small aircraft with high confort (much space between seats) on route A/D.

stackelbergIts competitor, Air Albi, later responded by differentiating and choosing big aircraft with little confort (little space between seats). In contrast to the standard Stackelberg result for undifferentiated goods, the first mover did not offer more seats than its follower, since the “most profitable customers” (the business passengers, who tend to pay less attention to price and more attention to comfort and to the convenience of departure times) are less numerous than leisure travelers.

Eventually, this maximum differentiation benefited both firms.


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

Loss Aversion, cheap capacity and investment in the Energy Game

In our last session of the Energy Game, players responded to a higher uncertainty about demand, and to the introduction of fossil-fuel power plants, with much higher production capacities:

Before the uncertainty increase, the average production capacity of a player was about 600 GWh/Rnd

Here are a few energy fleet the year after, following an increased uncertainty (same expectations about demand but with a higher variation coefficient: 0.3 instead of 0.15) and the introduction of fossil-fuel power plants.

EnergyMixHighUncertaintyOn most markets, total installed capacity exceeds 1400GWh/Round, even though the probability to have a shortage with that capacity is only about 0.6%.


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

Mergers and separate identities

In the last airECONsim session at ENAC, some of the teams decided to merge with robot airlines, and chose to keep a separate control over them.


This allowed them to build differentiated offers for customers. On this route, “Air Albi” operates flights with high comfort, while the acquired firm “robot 5 – Air Albi” operates cheap flights with very little space between seats.


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

CO2 Permits, Quotas, Grandfathering and switch to more ecological technologies

These graphs illustrate the evolution of prices, produced quantities and used technologies in the last session of “The Carbon Game”, at Ecole Polytechnique.

Each line corresponds to a different market with a specific environmental policy.

On “purple” markets, from year 2, fixed (non tradable) CO2 emission permits (quotas) were granted to the firms, and each firm had the same number of permits. On “green” markets, from year 2, emission permits were allocated in proportion to previous year’s sales (but the total amount of permits itself was fixed and independant of the firms’ actions). Banking permits from one year to the next was possible on green markets (grandfathering), but not on purple markets (fixed quotas). From year 4, the amount of quotas decreased by 20% every year.

On each market, firms could choose between 2 or 3 different technologies, some emitting less CO2 but being more costly.

The general trend on each market over the whole game was not very surprising, at least for the purple markets: Average prices tended to increase, productions tended to decrease, and more and more firms switched to the more ecological technologies as allocated permits got scarcer. More surprising maybe, the fact that nearly no firm ever used the option to bank permits on the purple markets (but it was used on other markets, involving auctions).

Also interesting and less intuitive,  in year 3, production soared and prices decreased on 2 markets. At this point, many of the firms switched at the same time to less polluting technologies, hence relaxing the capacity constraints created by the permits. And more production meant more competition in the price setting stage and eventually, lower prices (in spite of higher costs). This may seem a paradox, since less permits resulted here in lower prices paid by the consumers, but it was only a temporary phenomenon, due to discontinuous technology options.


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather

CO2 permits with badly designed allocation schemes

These pictures illustrate the evolution of CO2 emissions in one of the games of our last session at Mines Albi and Dauphine University.

Each line corresponds to a different market with specific environmental policies. The “light blue” line is a benchmark market over which there is no environmental policy. Without getting into details, from year 6, there are tradable emission permits on the three other markets and the cap (the total number of permits to be allocated) decreases by 80% every year (changes from one market to another bear on allocation specifications such as whether or nor banking is allowed, are the permits sold through auction or allocated for free, on the basis of past production or pollution, or else…)


In this first game, all the three policies seem to work pretty well. Still, something interesting is already visible in year 5 (the year just before the introduction of CO2 permits): emissions increase (more or less) in all three markets.

So, what happens in year 5?

Players are aware that permits are to be introduced the year after, and that the allocation in year 6 will be based on historical data (that + other info about the future of the markets):

  • On the brown and purple markets, players expect year 6 permits to be granted in proportion to their emissions for year 5.
  • On the red market, players expect year 6 permits to be granted in proportion to year 5 sales, but with a global and predefined cap.

So, it is no big surprise to see emissions increasing in year 5 on brown and purple markets. The increase on the red market does not come from a direct incentive to emit more CO2, but still occurs, as a side-product of the incentive to sell more (that, plus a correction of a mistake during year 4).

In fact, this is much more spectacular when you look at the other game of the session: Here the emissions in year 5 were so important that even in year 8 or 9, global emissions in the brown and purple markets were not so different from the benchmark level. The irony of the story is that the player who was the main contributor to this, found himself with many emissions permits, but with permits that were nearly worthless because of their number (later in the game, the permits were partly allocated on the basis of historical allocations and partly through auctions. And enough permits were granted through auctions to make them easily available to all the firms on the market)



In the next sessions, we will probably also introduce a “better-designed” allocation scheme. Emission trading schemes can perform quite well when they are well designed, but the devil lies in the details (in particular, because of expectations and governments’ commitment issues)


facebooktwittergoogle_plusredditpinterestlinkedintumblrmailby feather