Game Description

Buyers and sellers meet in a market—based on the double auction—for RobotDogs. The instructor chooses whether barking creates a negative externality (e.g., noise pollution) or a positive externality (e.g., crime reduction). You can then implement a correction for the market failure: tax or tradable permits for a negative externality, or subsidy for a positive externality.

Learning Objective 1: Externalities and Market Failure

The equilibrium of a competitive market without interventions for an externality-generating good is inefficient. There are too many transactions in the case of a negative externality, and too few in the case of positive.

Learning Objective 2: Pigovian Taxes and Subsidies

By reducing transactions, a tax can increase efficiency (total surplus) in a competitive market for a good generating a negative externality. By increasing transactions, a subsidy can increase surplus in a competitive market for a good generating a positive externality.

Learning Objective 3: Tradable Permits

By reducing transactions and ultimately allocating the right to consume to the highest willingness to pay, tradable permits can increase efficiency in a competitive market for a good generating a negative externality.

Brief Instructions

This game uses the same double auction as MobLab’s Competitive Market game. Students are randomly and evenly split into buyers and sellers. With Unit/Person > 1, buyers have decreasing marginal valuation and sellers increasing cost. A buyer can submit a Bid, while a seller can submit an Ask. A transaction occurs at the highest Bid if a seller presses Sell at Highest Bid or submits an Ask lower than the highest Bid. A transaction occurs at the lowest Ask if a buyer presses Buy at Lowest Ask or submits an Bid higher than the lowest Ask.

When a transaction is completed, a student can attempt to transact her next unit. In a market without government policy, a student's payoff for the round is her surplus from the units she transacts.

The default parameters (Group Size=16; Unit/Person=2, demand curve ranging from $7.40 to $2.90 and supply from $1.80 to $4.80) result in an equilibrium quantity of 12 (with no policy in place). By default, the Externality is negative in the Amount of $1.60 per transaction resulting in a reasonable difference between equilibrium and efficient quantities (9 for negative, 15 for positive).

As the nuisance or benefit from each RobotDog is split evenly amongst all Group Members, the default impact is $0.10 per person per dog transacted. This will generally be seen as sufficiently small and will not materially impact the decision to transact.

Especially with students experienced with MobLab’s double auction, you should see market prices to converge to the equilibrium prediction in the second round.

Finally, if you change the externality or the supply and demand curves in the Basic panel (or even Group Size or Unit/Person), click Supply/Demand Graph to view the resulting market.

Key Treatment Variations

Tax or Subsidy

Setting Policy to tax implements a per-unit tax equal to the marginal external cost, which in equilibrium results in the efficient quantity. Likewise, the per-unit subsidy equals the marginal external benefit. To make it clear that efficiency is enhanced, each student receives an equal portion of the taxes collected at the end of the round (or pays an equal portion of the subsidies distributed).


Setting Policy to Permits allocates tradable permits equal to the efficient quantity. A buyer needs a permit for each RobotDog she purchases. To make payoffs more equal (and to increase activity in the permit market), initial permit allocation is skewed to sellers with the highest cost and buyers with the lowest value. The permit market uses a double auction similar to the RobotDog market. Becuase of the complexity interacting with two markets at the same time (a RobotDog market and a permit market), it is advisable to increase Round Duration.


The following will help convergence to equilibrium

  1. Point out that each time there is a transaction, the price appears at the top of the list of most recent transactions on the trading screen.
  2. Make sure students are aware that a group's values and costs do not change between rounds. In this way students may infer that prices in the previous game may be relevant in the next game. Likewise, make sure to use the same distributions of costs and values when adding the tax, subsidy or tradable permits.
  3. Allow sufficient time for each trading period. For students without double-auction experience, start with a practice game with one three-minute round. With two items per player, two minutes per round will thereafter be more than sufficient (although you should allow additional time with permits).


Without any intervention, the market should converge to the inefficient equilibrium (which does not take into account the external cost or benefit from the externality). Use the Go To toggle between periods in a multi-period game. Click the Compare button to facilitate side-by-side comparison of two games within the same playlist. This is a great way to compare students' outcomes from a market with no intervention to one with a corrective intervention.

Table 1: Summary Table

We first present a table, which summarizes market outcomes by group (Figure 1) such as the number of transactions, average price, total surplus, buyer surplus, seller surplus, external surplus, and government surplus. We provide an average row to summarize these statistics across groups (markets). Efficient (socially optimal) and equilibrium benchmarks facilitate comparison of outcomes to predictions.

Figure 2: Supply, Demand, and Transactions

In each period, there are two charts for the selected group (Figure 2). On the left, we show the supply and demand graph. By default there is a divergence between marginal cost and marginal social cost by the size of per RobotDog external cost. On the right, we plot transaction prices (y-axis) across time (x-axis). The graph plotting transaction prices will contain at least one dotted line (two in the no-intervention case) to show the predicted price level. You can add/subtract this predicted price by using the checkbox in the legend. To change which group's results are displayed, click the radio button next to the group name in the Summary Table.

Robot Play

Our robot (i.e., an automated player) strategies for each role are the following (absent including tradable permits):

When tradable permits are included in the market, robot strategies are different. When the game begins, permits are distributed. Some buyers and some sellers will be given permits. This leads to four scenarios for robots, whose strategies are as follows:

tiled icons