If this book is supposed to give us a view of what is current and important in choice theory, it fails completely. It is hard to understand why it was included in Oxford's Very Short Introduction series, which is generally very good and sometimes remarkable. Just read Tim Gowers' book on Mathematics - enthralling!
First, its tone is inappropriate, in the sense that it states controversial things (remember, this is an introductory book) in a authoritarian manner. This style may go well with undergraduates, with whom unbalanced rants backed by very flimsy evidence will go unchallenged for fear of bad grades. For the rest of the world that does not depend on Mr. Allingham's goodwill and hopefully sunny disposition, it is singularly unpersuasive. It reads like robber-baron, Wall-Street/City of London capitalist propaganda, which is neither better nor more truthful than peasants-with-pitchforks Communist propaganda.
Second, it fails to take into consideration recent and not-so-recent advances in decision theory. Humans and animals are often irrational and have a very limited computational power. Thus models which assume rationality, perfect information and unlimited computational capacity are of little or no use. While perhaps interesting as intellectual exercises, they are useless in describing the real world. If you want to smell a rat, look in the pages of books like this - if it mainly deals with a "rational agent", any such book may be safely disregarded. The "rational agent" assumption, together with the assumption of gaussian probability distributions, have given the world a large number of horrible planning failures, from world wars to, most recently, the crippling of world economy by "talented" bankers, Mobutu-style kleptocrats who rake state-backed millions in bonuses after bankrupting everybody.
Third, something positive. While some of the author's assertions are wrong, they also are hilarious and thus provide good (if perhaps unintended) entertainment value. For example, the author states something to the effect that a certain amount of money means the same thing to someone who has a lot of money and to someone who does not. This assertion is so that it highlights his distaste for progressive taxation. So if you lose 10 dollars and you have 1,000,000, the loss causes the same pain as it would to a pauper who only has 100. Very convincing, isn't it? In reality, the perception of sensory stimuli and numerical quantities is ruled by the Weber-Fechner law, which states that quantities are perceived relative to the baseline and for which overwhelming experimental and observational evidence exists. In other words, a loss of 10 dollars for the 100 dollars pauper is approximated by a loss of 100,000 for the millionaire. Yet you wouldn't learn it from this book. The Weber-Fechner law has been formulated in the 19th century - who knows, news may travel slowly from Germany to England.
For decision theory books that actually do apply to real decision-making, it is much better to read Daniel Kahneman & Amos Tversky, Howard Raiffa & John Pratt & Robert Schlaifer, Paul Glimcher, Sheena Iyengar, Dan Ariely, Barry Schawartz, Jonah Lehrer. These books have various lengths, areas of emphasis and levels of mathematical treatment, yet none falls into the category of this introduction. Hopefully they will satisfy your needs and intellectual curiosity. This book has no place alongside them.