In the Somali city of Mogadishu, in a few short hours in 1993, the US military suffered a deadly, and very public, defeat. Six in ten of the army rangers despatched to track down a local warlord never returned. Their bodies were dragged through the streets in front of cheering Somalis and the world's television cameras. President Clinton ordered a prompt, humiliating withdrawal. The images faded from the public's mind, but US foreign policy wonks didn't forget, and started trying to figure out what went wrong. The conclusion they came to was that America needed to prepare for a 21st century of "urban slum combat".
It is in this apocalyptic vein that Mike Davis concludes Planet of Slums. He notes that "the Pentagon's best minds have gone where the United Nations, World Bank or Department of State fear to go". The conditions of today's global slums are so wretched that they can no longer be seen as simply a moral calamity; they represent a threat to global security. Davis starts from another, equally alarming, premise. In the near future, there will be cities of unprecedented size, with unprecedented numbers of people packed into them. Today only Tokyo has a population of 20 million; two decades from now, at least ten Asian cities will.
The author deploys a range of neologisms in depicting a world of "metropolises", "hypercities", "megacities" and "megalopolises". But the gist of his argument is straightforward: the slums of the future will be huge, and will contain people living in staggering poverty. He paints a picture of residents living without basic amenities, suffering underemployment, crime and disease; of governments periodically clearing - or "beautifying" - their makeshift homes, often to make way for middle-class housing. Such descriptions of never-ending misery might almost seem comic were they not so tragic.
Meanwhile, the IMF and World Bank hobble developing countries with austere international macroeconomic policies, while seemingly friendly NGOs feather their own nests. And as if that weren't enough, such countries are disproportionately more likely to suffer earthquakes, tsunamis and pollution.
Davis's descriptions of the conditions endured by slum-dwellers provide reason enough to read this book. His analysis is full of gripping stories from globalisation's front line - such as that of wealthy citizens in developing countries escaping the squalor surroun-ding them by building pseudo-Californian gated suburbs. Nevertheless, Planet of Slums is an oddly unsatisfying book. Part of the problem is Davis's infuriat-ing method of arguing. Having claimed that some thorny problem - pollution, say - is endemic across all slums, he will then pick a few graphic examples as evidence. But this proves neither how severe the problem is generally, nor if things are worse in one place than another.
Such hopscotching is part of a larger problem of discrimination. Davis seems unable (or unwilling) to identify any positive development, useful reform or potential way forward. He finally admits, 160 pages in, that many experts regard Indian and Chinese growth as a qualified success story. Yet even when confronted with grounds for optimism, Davis can see little but immiseration and injustice. He never properly explains the millions lifted from poverty, the burgeoning Asian middle class or the huge progress made in cities in South Korea, Thailand and Malaysia. The situation is grim, but not as hopeless as Davis would have us believe. Not every aid regime is ineffective; some developing countries deal with slums better than others. Ultimately, Davis is so keen to find horror in every city that he sees hopelessness everywhere.
The same cannot be said of another book about urban poverty, written a century and a half ago: Friedrich Engels's The Condition of the Working Class in England. Engels wanted to horrify polite society, but also to suggest a way forward: to prove that slums were not inevitable, that human beings need not live in such conditions. Davis, for all his powers of criticism, stops short at the point where we need him most.