RDA's spend around £1.5bn per year. Administration accounts for about 7% of this. The Homes and Communities Agency is budgeted to spend £6bn this financial year. Administration accounts for less than 2%. It should be clear that efficiency savings in the delivery (e.g. through abolishing these quangos) will only make a small difference to the overall expenditure. The new government is going to have to cut programme expenditure. So the crucial question is: what should it cut?
I would significantly decrease government expenditure on the provision of commercial real estate. In a few places (the centres of Manchester, Leeds, Newcastle) such expenditure has probably positively reinforced developments in the private sector. In poorer areas and cities with weaker economies, I believe that shiny new government funded buildings simply transfer employment from other places in the immediate area and as a result don't provide significant employment benefit.
I would continue some of the public realm expenditure in more deprived neighbourhoods. I would focus on the public good (amenity) benefits these deliver to poorer families. For example, public space provision (e.g. parks) likely has large amenity benefits in poorer neighbourhoods where families have little private space. The benefits of signature buildings and public art are, shall we say, less clearly defined.
Improving the quality of private space (e.g. via decent homes) clearly benefits poorer families, although it may cause other problems - e.g. decreased incentives to work. That said, crappy social housing is surely not the best way to deal with these problems so some component of decent homes expenditure should stay.
The government should spend less on building houses. According to the HCA around a half of the homes delivered this year will have been (part) funded by the government. This figure is distorted by the recession but why does the government need to be so involved in the provision of affordable housing? The simplistic answer is that "house prices are too high". But what drives this are policies which raise costs: e.g. the national brownfield building target (because building on polluted land is expensive) and supply restrictions from the planning system. In effect, government restrictions raise land costs, which raise the price of housing, which makes housing unaffordable, so the government spends large amounts of money to provide affordable housing. Government has it in its power to change these restrictions (e.g. through taxing developers and land owners and giving the money to local homeowners, LAs etc, negatively affected by new building).
Moving away from the built environment, there is a case to be made for reducing the amount spent on corporate welfare (specifically giving money to business to locate in poorer areas). I would stop funding pretty well everything which simply shuffles around the existing pool of high skilled workers and focus instead on things that directly improve the employment prospects of poorer residents. Generally, I think that corporate welfare is an effective way of getting firms to locate in places that they otherwise wouldn't. But it is expensive and it doesn't turn areas around (at least on the basis of fifty years expenditure to date).
If the cuts fall in all these areas, it would allow us to protect expenditure that I think matters most - that on improving educational outcomes of poorer children and improving the labour market prospects of poorer adults.