We have already discussed one common metric for measuring the exercise of market power (what we had called "behavioral market power" earlier in the lesson) - the Lerner Index. Now we'll turn to how structural market power is typically measured, and discuss how a measure of market power used widely across many industries is not suitable for electricity markets.
Typically, a firm will possess structural market power if that market is concentrated, meaning that there are very few firms, or if one firm has a very large market share. One of the most widely used metrics for market concentration is the Herfindahl-Hirschman Index (HHI). The HHI is defined based on the sum of the squared market shares of all firms in the relevant market:
In the HHI equation, Si represents the market share of the i-th firm, written as a whole number rather than a decimal. For example, if Firm M has a 10% market share, we would write that as SM = 10 and not as SM = 0.1.
As a simple example of how to calculate the HHI, suppose that there were two firms in the market, one with a 60% market share and one with a 40% market share. The resulting HHI for this market would be:
The HHI is bounded from above by 10,000 and from below by zero. This is also easy to see - if a market has one firm (a monopolist) then the HHI would be 1002 = 10,000. If a market had a huge number of firms, each of which had a very very small market share (much less than 1%) then the HHI would be very, very small.
It is sometimes helpful to keep in mind what the HHI would look like for markets with certain numbers of equally sized firms.
- A market with two equally sized firms would have an HHI of 5,000
- A market with three equally sized firms would have an HHI of 3,333
- A market with four equally sized firms would have an HHI of 2,500
- A market with five equally sized firms would have an HHI of 2,000
- A market with six equally sized firms would have an HHI of approximately 1,800
As a rule of thumb, markets with an HHI of 2,000 or greater are judged to be highly concentrated. Proposed mergers that would increase the HHI much beyond 2,000 generally get a high level of regulatory scrutiny, on the grounds that such a concentrated market would harm competition.
While the HHI is used very commonly (and if you read the State of the Market reports from the various IMMs, you will see the HHI used), it is of limited usefulness for electricity markets. Let's look at the HHI for California's failed electricity market as an example. The table below shows the market shares and their squares for the largest firms in California's electricity market.
|Firm||Market Share (%)
|Pacific Gas and Electric||10.5 (110)|
|City of Los Angeles||8.4 (71)|
|Mirant (Southern)||8.0 (64)|
|NRG Energy||7.4 (55)|
|Duke Energy||7.3 (53)|
|Southern California Edison||6.2 (38)|
|Others (cumulative)||30.1 (0)|
|Implied HHI||632 (10,000/632=
15.8 equal sized firms)
If we were to calculate the HHI for California's electricity market, we would get 632, as shown in the table. This is roughly equivalent to a market with around sixteen equally sized firms - hardly the stuff that would make competition regulators nervous. But we have already seen that California's market was highly susceptible to manipulation. How is this possible if the market is really that unconcentrated?
As it turns out, there are two problems with the HHI when applied to electricity markets. The first problem is that electricity cannot be stored at large quantities at reasonable cost. The second problem is that power grids have some public good characteristics and there are large network externalities. As California found out the hard way, if suppliers try to act anticompetitively, then blackouts or other severe consequences can result. During periods of peak demand, manipulative actions by small suppliers can have large impacts on the market or on the power grid itself. This is impossible to tell just by looking at the HHI.
Let's take a specific numerical example. Suppose that total capacity in an electricity market was 100 MW. Demand is 90 MW. One firm, Firm M, owns 18 MW. The rest of the capacity in the market is owned by 82 small suppliers who own 1 MW each. You can verify for yourself that the HHI in this market would be 324, which would indicate that the market was unconcentrated and should behave competitively.
Everyone in the electricity market, however, knows that demand is 90 MW. In particular, Firm M realizes that it is in a very strong strategic position. If Firm M withholds its capacity from the market, then the RTO will not be able to meet total electricity demand, because it would have only 82 MW of capacity to dispatch for a demand of 90. Since the RTO must buy from Firm M in order to avoid blackouts, Firm M actually has monopoly power in this situation. Firm M has this monopoly power even though Firm M is not that large relative to the rest of the market, because the capacity controlled by Firm M is larger than the surplus capacity in the system (total system capacity - total market demand).
Because Firm M effectively has a monopoly position in the market, even though the market as a whole is not concentrated, Firm M is a special kind of monopolist. We refer to Firm M in this situation as a "pivotal supplier" or more properly a "pivotal monopolist." Just as Firm M found itself in a situation where it had pivotal monopoly power, it is possible for multiple firms to jointly be pivotal suppliers. Let's work through a second example to illustrate this. The electricity market again has capacity of 100 MW and demand is 90 MW. Firm M controls 8 MW and Firm N controls 8 MW. The remaining 84 MW of capacity are controlled by small firms that own 1 MW each. Once again, you could verify for yourself that this is an unconcentrated market (the HHI would be 212).
The RTO, however, must purchase capacity from both Firm M and Firm N in order to avoid blackouts. If Firm M and Firm N both decided to withhold their capacity, then the RTO would be unable to meet demand. In this case, Firm M and Firm N would be jointly pivotal suppliers. The two suppliers together would form a pivotal duopoly and would be able to jointly control the market price if they wanted to do so.
We are now in a position to define a pivotal supplier or group of pivotal suppliers more generally. This definition can be used as a test or "screen" to determine if a given firm or group of firms is a pivotal monopolist or part of a pivotal group.
First we'll develop a definition for a pivotal monopolist. Let KM be the capacity controlled by a single firm (Firm M), let Ktotal be the total system capacity, and let D be demand during some period. Firm M would be considered a pivotal monopolist if:
If this condition holds, then the RTO needs to use capacity controlled by Firm M in order to meet total electricity demand and Firm M would be a pivotal monopolist.
Next we'll extend this definition to define a pivotal group consisting of n firms (so n = 2 would correspond to a pivotal duopoly, for example).
Let represent the total capacity of the n firms that may form a pivotal group. Typically Kn would include the capacity of the n largest firms in the market. Ktotal and D are defined as in the pivotal monopolist case. Then the group of n firms would be considered jointly pivotal if:
Before we go back into some more examples of pivotal supplier calculations, let's reflect for a minute on some of the major differences between the HHI and the pivotal supplier test as indicators of structural market power.
The first difference is that the HHI is a measure of overall market concentration, whereas the pivotal supplier test is a measure of the concentration of the system surplus capacity - that difference Ktotal - D between the total system supply and demand at any particular time period.
The second difference is that HHI changes only when there is new entry into the market, exit from the market, or if there are ownership changes that affect market concentration. The pivotal supplier measure changes as market demand changes. The higher are system demands, the more likely it is that a single supplier or group of suppliers will be pivotal.
Thus, the pivotal supplier test needs to be run periodically for the RTO to get a sense of whether there are groups of suppliers who are able to manipulate market prices. There is no hard and fast rule as to how frequently the pivotal supplier test is run. PJM is probably the most aggressive RTO; it runs a pivotal supplier test every five minutes to identify pivotal monopolists, duopolists, or groups of three jointly pivotal firms. This is called the "Three Pivotal Supplier Test" and you can read more about it in the State of the Market Report for PJM. Other RTOs run their pivotal supplier tests less frequently and generally under peak demand conditions.
Finally, we'll take a look at how the pivotal supplier test would be executed, using the California electricity market from Table 11.1. The table is reprinted below, and the capacity ownership has been scaled so that Ktotal = 100 MW.
|Pacific Gas and Electric||10.5|
|City of Los Angeles||8.4|
|Southern California Edison||6.3|
Suppose that demand during some hour was 90 MW. Is there a single pivotal supplier during this hour?
To test for a single pivotal supplier, we would compare the capacity of the largest supplier (AES) to the spare capacity in the system. Here we note that KAES = 12.5 MW, Ktotal = 100 MW and D = 90 MW. Directly applying the pivotal supplier test equation, we see that 12.5 > (100 - 90), so AES would be considered a pivotal monopolist.
Suppose that demand is now 80 MW. Is there a single pivotal supplier during this hour?
To test for a single pivotal supplier, we would compare the capacity of the largest supplier (AES) to the spare capacity in the system. Here we note that KAES = 12.5 MW, Ktotal = 100 MW and D = 80 MW. Directly applying the pivotal supplier test equation, we see that 12.5 < (100 - 80), so AES would not be considered a pivotal monopolist.
The two largest suppliers together, however (AES and Pacific Gas and Electric), control 23 MW. If we apply the pivotal supplier test for AES and Pacific Gas and Electric jointly, we would see that 23 > (100 - 80), so AES and Pacific Gas and Electric jointly would be considered a pivotal group.
Here is an example that you can try on your own:
Suppose that demand was 75 MW. Show that AES is not a pivotal monopolist; AES and Pacific Gas and Electric jointly are not a two-firm pivotal group; but AES, Pacific Gas and Electric, and Reliant would jointly be a three-firm pivotal group.