1996–1999Card 3 of 4

Market Design Debates

Pool vs. bilateral: a battle of philosophies.

Two Visions of the Market

As ISOs formed across the country in the late 1990s, a fierce intellectual battle erupted over how wholesale electricity markets should be designed. The debate had two poles.

The centralized pool model: All generators submit price offers into a central market. The ISO accepts offers in merit order (cheapest first) until demand is met. All accepted generators receive the same "clearing price" — the price of the last unit accepted. This model maximized transparency and enabled sophisticated products like financial transmission rights.

The bilateral trading model: Generators and load-serving entities negotiate contracts directly, one-on-one, like any other commodity market. The ISO operates the physical grid and handles the last-mile balancing, but does not set prices. Price discovery happens through thousands of private negotiations.

The British Experiment

Great Britain, which had restructured its electricity market in 1990 — ahead of America — had tried a version of the centralized pool model called the "Pool." The British Pool had significant problems: generators could exercise market power by withholding capacity to drive up the clearing price. The Pool was eventually scrapped in 2001 in favor of bilateral trading plus a balancing market.

American market designers watched the British experience closely and drew different lessons. Some concluded that centralized pools were inherently flawed. Others concluded that the British Pool's specific design features — mandatory participation, no demand bids — were the problem, not the pool concept itself.

PJM Wins the Design Battle

In the long run, the American market converged on a hybrid model closer to PJM's centralized design than California's or Britain's experiments. The key innovations:

  • Locational Marginal Pricing to handle transmission congestion efficiently
  • Day-ahead markets that give generators and load a financially binding schedule
  • Real-time markets that balance deviations from day-ahead schedules
  • Capacity markets to ensure long-run adequacy (addressed in Chapter 4)
  • Ancillary services markets for reserves, frequency regulation, and voltage support

This design, refined over two decades, now underlies most organized wholesale markets in the United States. It's complex, imperfect, and perpetually under revision — but it has proven far more stable than the alternatives.

Vignette

London, March 27, 2001. Britain's Electricity Pool — the world's first mandatory centralized power market, launched in 1990 — was shut down and replaced with the New Electricity Trading Arrangements. The Pool had become a textbook case of market manipulation: generators submitted bids, a central dispatcher selected the cheapest combination, and everyone cleared at the single highest accepted price. Large generators learned to withdraw capacity and submit inflated bids to move the clearing price upward. NETA replaced the centralized pool with a bilateral trading system, where generators and suppliers contracted directly and only imbalances settled through a central mechanism. Within months of NETA's launch, wholesale prices fell 25 to 30 percent. American RTO designers absorbed two lessons: centralized pools needed strong market power mitigation, but bilateral markets raised their own concerns about price transparency. The debate shaped every U.S. market rule written in the following decade.

Ofgem — NETA Review, August 2001 ScienceDirect — "Failing Electricity Markets: Should We Shoot the Pools?"

Ask about: Market Design Debates