EELING hungry? How about a peanut butter and jelly sandwich? But be careful how you make it - you don't want to infringe on United States Patent No. 6,004,596, for a "sealed crustless sandwich."

Well, maybe it is safer just to have some toast. But watch out for Patent No. 6,080,436 on a "bread refreshing method."

Not hungry anymore? Well, maybe you could just play with the cat. But even that's not safe. Patent No. 5,443,036, "Method of Exercising a Cat," protects using a laser pointer to amuse a cat "and any other animals with the chase instinct."

These patents, which might strike some as frivolous, are symptoms of the deeper problems with the nation's patent system discussed in "Innovation and Its Discontents," a recent book by two economics professors, Adam B. Jaffe of Brandeis and Josh Lerner of Harvard.

Most economists, including the authors of this work, accept the need for intellectual property protection. Patents provide necessary incentives for innovation by adding "the fuel of interest to the fire of genius," in the words of Abraham Lincoln, the only American president to hold a patent - for a device to lift boats over shoals.

However, any system of intellectual property protection requires balance. The law must offer enough rights to the inventor to stimulate innovation, while at the same time ensuring that the resulting temporary monopoly on the invention does not impose too onerous a burden on society.

A patent has a length, a width and a height: how long it is valid, how broadly it applies and how novel it is. Mr. Jaffe and Mr. Lerner are most concerned with the last issue.

A patentable invention must be "not obvious to one skilled in the art." This guideline is subject to interpretation, and in recent years it has seemed as if almost anything could be patented.

According to the authors, the deterioration in patent quality started in 1982, when Congress made two seemingly innocuous changes to the system.

First, it created a specialized Court of Appeals for the Federal Circuit. The motivation for this change was well intentioned: the district courts that heard patent appeals before 1982 applied widely varying standards, and the new Court of Appeals was supposed to offer a consistent policy.

Second, Congress tried to make the processing of patents self-sustaining: it was to be supported entirely from fees paid by patent applicants.

These reforms resulted in a tilt toward patent holders. Mr. Jaffe and Mr. Lerner assert that since these changes, a patent applicant is "much more likely to have the patent granted; the patent is much more likely to be held valid if challenged in court; and the party accused of violating the patent is more likely to be found to be an infringer and forced to pay a large monetary award."

The economists explain these changes as an example of what economists call regulatory capture. Regulators - in this case, judges who hear patent cases - are often drawn from the industry they regulate. Even in cases where they come in as outsiders, over time they tend to identify with the concerns of the industry they regulate rather than with the public at large.

The Patent and Trademark Office now views its constituency as patent holders - not surprising, given that they make up the group that provides its financing. As one patent examiner put it: "When I first started here, I was told, 'When in doubt, reject' and to try to reject. Now I am told, 'When in doubt, allow' and try to find a reason to allow."

Patent examiners are underpaid and overworked. Novelty is supposed to be judged relative to "prior art," but it is almost impossible for a single human being to root out all relevant prior art. One result has been a steady reduction in patent quality, with patents of dubious novelty being granted routinely.

The reduction in standards for granting patents has started an arms race among patent holders. The best defense against being sued for patent infringement is to hold a portfolio of patents that can be used to countersue. So every company has to build a "patent thicket": an arsenal of patents that can be used offensively or defensively, as necessary.

The result has been sharp growth in patents and litigation. When patent suits are brought, they can be horrendously expensive. One study found that in cases where at least $25 million was at risk, the cost of defending a patent infringement suit was $2 million to $4.5 million. Even with small cases involving less than $1 million, the cost of defending was nearly half the amount at risk.

So what is to be done?

The first set of reforms that Mr. Jaffe and Mr. Lerner propose is to create better incentives to discover prior art. Once an examiner decides a patent might be granted, the Patent and Trademark Office would issue public notice of an "intent to issue," and concerned parties could offer information that could be relevant to the determination of novelty.

Such a procedure is not novel - European countries have used such patent opposition procedures for years. A 1992 American commission recommended such a procedure here, and a relatively weak form of patent opposition has been allowed. Mr. Jaffe and Mr. Lerner offer several detailed suggestions for strengthening this tool.

The second reform they propose is to offer several escalating levels of patent review, so that trivial patents could be dealt with at lower levels, reserving detailed examination for important cases.

The third reform is to replace jury trials with judges, who could hire special masters, experts who would work for the judge in examining the arcane technological debates that arise in patent cases.

They also offer several other suggestions, like moving from the current "first to invent" rule to the "first to file" rule. This would not only align the United States with the rule used in the rest of the world, but would also eliminate legal wrangling about priority.

Mr. Jaffe's and Mr. Lerner's book offers a lucid, entertaining and sobering look at the American patent system. Their suggestions are worth serious consideration by policy makers and legislators.

Hal R. Varian is a professor of business, economics and information management at the University of California, Berkeley.