De Soto explains that there is a lot of property in the 3rd world that economist's don't normally count. It is all the unofficial, or "extralegal" property. Many of the people, for example, are squatters living in shanties they have built in slums, working at unofficial businesses. The value of their land, buildings, and businesses is often a majority of the total property of a 3rd world country.
People choose to keep their property outside the law because complying with the law and making their property holdings and businesses legal is prohibitively difficult, taking many months or years of work, with hundreds of bureaucratic steps involved.
All this informal property is very hard to deal with -- you cannot use it as collateral for a loan, for example. Building a shanty in a slum is very inefficient, you have to be living in it the whole time to enforce your stake in it, building with what materials become available. If the land had been formally purchased, one could take out a loan against it to fund the construction. Because the state doesn't recognize your property, you often turn to gangsters and the like to guarantee your property rights.
De Soto, a Peruvian, provides many examples from Peru and other 3rd world countries. He spends a lot of time discussing American history -- in much of the nineteenth century, many American settlers and miners had extremely dubious claims to the land they settled and worked, and he discusses how this property eventually evolved into formal, legally recognized property.
Then the author discusses how this process could be brought about in the modern third world. The barriers are formidable, as many of the elite classes and particularly the lawyers are quite unsympathetic to the plight of those who own unofficial property.
De Soto makes a convincing case for a process that is central to spreading prosperity to the third world. The book has many endorsements from prominent people and prestigious economic publications.The main problem with this book is that it repeats itself too much. It could have been about 1/4 as long as its length of 246 pages.