Good news - Matt Yglesias is to write a book about housing policy, provisionally entitled 'Rent is Too Damn High' (see also). I'm fairly sympathetic to the broad Yglesian position on housing policy, which I understand to include the following points:
Pareto improvements are extremely thin on the ground in housing policy, more so than in most areas. This means that it is very difficult to please one group of people without severely displeasing another. For example, loose housing supply in a local area benefits newcomers but harms existing owner-occupiers, while tight supply harms (by excluding) newcomers but benefits existing owners. Existing owners tend to be economically self-interested and more politically powerful than would-be newcomers, so supply tends to be tight. Also, fragmenting land ownership arguably creates a kind of 'anti-commons' wherein individuals can block changes that would benefit society as a whole.
Housing supply decisions have economic, environmental, aesthetic and social ramifications, often in different directions - but very few of the actors involved in debates over housing have any incentive to take all of these into account, let alone the ability to do so. For example, a block of flats in a nice area might be undesirable on aesthetic grounds but provide cheap accommodation for those who need it. There are real and important dilemmas here, but they often seem to go unacknowledged. Another way of saying it is that housing policy affects the distribution of many important resources as much as income does, yet its distributional role seems to go mostly undiscussed.