Another hoary myth about economic growth has arisen. This one says that the floods in the mainland have been very bad indeed but, on the bright side of things, there will now be a big effort to repair the damage and all the spending will count as investment, which will help boost economic growth. The fallacy lies not so much in the argument as in the peculiar way that national accounts are calculated. Gross domestic product is a concept that was cobbled together in the 1920s and precious little has been done to modernise it. Governments may be grateful that they don't have to let accountants anywhere near it, as it would not stand up to a corporate audit longer than two seconds. For one thing, it treats construction and destruction as equally productive. Take a car-park and build slum housing on it. This counts as fixed-capital formation. Five years later, when the depressed tenants have trashed the place, tear it down and turn it into a car-park again. This also counts as fixed-capital formation and, in the process, a new branch of mathematics has been discovered. One minus one equals two. The concept of amortisation has simply been evaded altogether. This is not tolerated in corporate accounts. If a company builds a dike to hold back potential flood waters from its plant, the amount spent on building the dike goes into its balance sheet as investment in fixed assets. If for some reason the engineers then say that the dike will last for only 25 years until it needs to be replaced the company will write down the value of that dike by 4 per cent each year in the balance sheet and deduct the same amount from earnings each year. At the end of the 25 years the dike must be rebuilt but its value has also gradually been written down in the accounts and is now carried in the balance sheet at zero. Everything has balanced out. When the mainland tots up its national accounts, however, the cost of building a dike such as this is shown as fixed-capital formation in the year it was built, then vanishes from the gross domestic product figures in subsequent years. When it then breaks in a flood, there is no write-down either of the value of the dike or of the plant which it protected and no deduction from fixed-capital formation for the loss. This is the only way that restoration efforts can be made to appear as an increase in investment and a boost to the growth rate of gross domestic product. But it is an illusion. National accounts are not a balance sheet. They are not a profit and loss account. They are a truncated form of cash-flow statement and subject to an enormous range of error, where corporate cash-flow statements are precise. The harm this does goes further than just deluding a public, which treats them as a form of economic report card. Governments increasingly build their reputations on achieving a high GDP growth rate and in many countries are prepared to ruin national assets if this has no effect on GDP figures while development will push the figures up. They would think twice about it if cutting down forests meant both writing down the value of those forests in the national accounts and providing for losses that the subsequent erosion and floods are likely to bring. It could mean that these acts of destruction would bring the GDP figures down further than later spending on the development of the resulting wasteland would bring the figures up again. If that were fully shown in the national accounts, the forests might be saved. Various bodies of economists have from time to time suggested doing this, but the inherent problems are immense. What value does one set on lands, forests and other forms of national heritage? How does one overcome the lobbying efforts of businesses that benefit directly from development but lose only indirectly on destruction. How does one convince poverty-stricken people to deprive themselves even more? Good questions, but they don't help overcome the failings of national accounts.