I do wonder how good the data on investment is, and whether as the service sector grows it is systematically underestimating investment. How well is investment that essentially takes the form of hiring labour to work on intangible assets, or simply to constitute an accumulated hired factor production (human capital), accounted for in all this data that tells us investment has been falling?
Posted by: Luis Enrique | March 11, 2010 at 05:24 PMThis seems a pretty reasonable hypothesis. Any idea how it could be tested? The last thing that those service businesses want is for their intangible work to count as capital investment - they'll have to pay corporation tax on it until they can write it off in four years. So some firms may not want to report investment figures to the government.
However, at least one form of non-tangible investment - scientific R&D - brings tax benefits in some countries, and companies do have an incentive to report this.
Nick Rowe pointed out a while ago the difficulty of directly measuring investment, by consumers or businesses. And there are many grey areas - what if I have a new brochure designed which I plan to keep using over the next three years? This is a one-off cost, much of the benefit of which will accrue in future years; but certainly some of it is a current-year revenue expense. This would normally be written off as a marketing expense but you could make a good argument to count some of it as an investment. If the design is a good one, the company's accumulated intangible assets and goodwill will certainly be worth more at the end of the year than the beginning.
My thought is that, mathematically, statistically, or from an accounting viewpoint, it should be possible to derive an investment figure. The only numerical way I can think of to do this is to split out cost of sales from overheads and try to look at forward estimates of earnings for the next year. But both of those figures are really hard to estimate accurately and I would not want to rely on the numbers submitted by firms in their official accounts. Perhaps it would have to be done by surveys or even detailed interviews with a representative sample of firms; after all that's good enough for the employment figures, why wouldn't it work for investment too?