Imagine that Chancellor Alistair Darling had decided to boost public spending by investing £20 billion of taxpayers' money over the next two years in buying land and building homes earmarked for eventual open market sale.

What would be the net result?

Well we would have 200,000 more homes to help meet the much talked of housing crisis. We would save about 150,000 construction jobs plus many others in associated businesses.

And what is really tasty is that the deal would pay back all that was borrowed and could earn the Treasury an extra £16 billion, which in turn it could use to reduce public debt.

It sounds too good to be true. But if you make reasonable assumptions (see below) it seems that this is what the numbers tell you. And what is more it appears very low risk.

Here is where I started.

The outline idea is for the Government to fund, through an arms-length private market focused body, the development of 200,000 homes over two years and then to sell them again when the market price recovers, say in 2014.

To get a handle of the cost breakdowns I went to the presentations to analysts made by Bovis, Persimmon and Barratt. These together provided some detail on the typical cost elements of a new-build home. From this I estimated a rough cost model, which I checked against the sort of numbers provided in the Callcutt review.

I took the average selling price of a new home to be £180,000 up to the end of 2007. That is just below peak. I called the gross margin 20% (that is a pre-collapse margin) which put total costs at about £144,000.

The elements will vary place to place, product to product, but an average breakdown looks something like:

  • £39,000 - Land
  • £38,000 - Labour
  • £55,000 - Materials
  • £12,000 - Sales & Marketing and other costs

If the build rate is 100,000 a year that would require about 150,000 people engaged in the house building process.

So taking one year's output of 100,000 homes. Here are the sums.

Cost to build 100,000 homes for sale at 2007 prices

£14.4 billion

Savings to the Treasury

£0.95 billion: Revenue to the Treasury from the tax taken on labour

£1.20 billion: Savings to the Treasury from reduced benefits of 150,000 fewer unemployed

£1.30 billion: From buying the land at currently reduced prices (a third discount)

£0.95 billion: Reduction in sales & marketing and other costs

Net cost to the Treasury to build 100,000 homes

£10.0 billion

Market price at which to sell

£18.0 billion

To cover the spending over two years would mean borrowing £20 billion, which would attract interest at, say, 5% of £1 billion a year.

That means to cover the period up to when the homes are sold each home would have to generate £5,000 a year in rent, net of other expenses, such as maintenance, repairs, improvements and administration. This is easily achievable.

When the local house prices hit a set target the Government can sell each of the 200,000 homes off at, say, an average of £180,000 returning £36 billion to the Treasury, making a return to the taxpayer of £16 billion. And the Government could waive stamp duty into the bargain.

Naturally life isn't that simple, there are many clouding factors - not least political - and other influences will impact on the real outcome. You can not expect to pull 150,000 people into the production of homes without other effects, both negative and positive.

But here are a few of the missing effects on the plus side I have not taken into account in the simple calculations above.