We like our invention mythology in the UK. Our favourite version goes: genius inventor produces idea (the most difficult and creative aspect of innovation); inventor patents idea, creates economic value and technological superiority for the UK; and there we are, all the way to a knowledge economy.

The common myth in this socially wholesome illusion is the relative importance of the idea versus implementation, but James Bessen and Michael J. Meurer explode a few more illusions in their hard-hitting analysis of how patents perform economically.

We hear that small inventors gain less value from patents than large corporations; justice is obscured by patent trolls and patent thickets that exploit the abstract technologies such as software; researchers take out early-stage patents in complex technologies such as biotechnology that ultimately disincentivise innovation. The authors challenge our seductive assumption that patents promote economic growth by providing incentives to inventors, their employers and investors.

What is the significance of the word "property" in intellectual property? Economists and lawyers equate intellectual with physical property and this, say Bessen and Meurer, is a problem. We usually draw property rights with fences, rules of ownership and the penalties for trespassers clearly marked. Thus owners understand their rights and can buy, sell and divide with certainty - and contribute to economic growth.

ÁñÁ«ÊÓƵ

Drawing an idea - and let's take the cartoon depiction for simplicity - is usually done as a thought bubble; a fluid, cloud-like entity with fuzzy boundaries.

Mapping the boundaries of a technology, such as the extraction of umbilical cord blood, is much more complicated than a plot of land; even if potential innovators try to avoid infringement, the scope and validity of the patent can be uncertain and can change over time.

ÁñÁ«ÊÓƵ

Where you can draw a boundary, for example, around the small well-defined molecules of the chemicals and pharmaceutical industries, patents work better. But in the abstract technologies, notably software, litigation is predictably soaring.

The book claims that patents, unlike property rights, have only a "weak and indirect relationship to economic growth". Its empirical evidence suggests that the majority of patents arise in industries that don't innovate much, that the net incentives after litigation costs and protection are smaller and that patents contribute less to technological leadership than we think.

This is perhaps arguable, and a US-centric argument at that, but it includes a valid point that net patent incentives need to be large enough to be effective policy instruments and therefore should be judged relative to other policy instruments that can be used to increase investments in research and development. Other models such as collective research and development and open innovation - characterised by more diversity and openness between innovators - appear as well suited to genuinely breakthrough technologies as are traditional ones.

Is this failure though? Patents do offer rewards for small innovators, for chemical and pharmaceutical companies, for lawyers and indeed universities. Complex technologies make boundary definition difficult and interpretation in patent law is as hard as any area of law. It's probably possible to make an equally elegant, and opposite, argument that patents work economically. Bessen and Meurer state that "only modest reforms are politically feasible". That's certainly true.

ÁñÁ«ÊÓƵ

Their aim in writing the book was to move beyond "faith-based policy" and base patent policy on the evidence. In this respect, this is an important book, for policymakers, lawyers, scholars and also for universities. Under the Government's third-stream strategies, universities are seen as a set of intellectual assets, the idea being that transferring these assets, under various knowledge-transfer mechanisms such as technology transfer, will produce social and economic benefits. For the past seven years, research-intensive universities' knowledge-transfer ability, ergo economic benefit, has been measured largely by the amount of intellectual property it produces. The authors' empirical evidence that challenges this causality should be explored.