Abstract

Depreciation has been and remains a source of confusion and dispute amongst economists. Failure to use clear, valid and self-consistent definitions for output (GDP), depreciation (δ) and income (NDP) is one major cause.

Definitions of depreciation are varied and imprecise with obsolescence being the exception. It is usually assumed, without evidence, to be the most important. It is, however, measurable but it has not, as far as I can find, previously been measured. I show that it accounts for only about one third of total US depreciation, and I expect this result to surprise economists.

We have no data from which we can identify the causes of depreciation which do not result from obsolescence. However, in the absence of any other cause which can reasonably be assumed to have had such a major impact, climate change is, in my judgement, likely to be the single most important cause of depreciation and even more important than obsolescence.

Investment is not just needed for growth. Historically the US has needed to invest between 3.5% and 4.5% of its national income (NDP) in new capital simply to prevent living standards from falling.

Standard models assume that growth is determined by changes in technology and additions to the supply of labour and capital. This assumption is rendered invalid by the evidence of the importance of non-obsolescent depreciation.

A major revision to US national accounting was made by the Bureau of Economic Analysis in 2013 by altering the definition of output. This was a mistake which should be rescinded.

The errors made by the advocates of this change included a bad definition of income, which they conflated with rather than distinguishing it from depreciation, which is not part of income but is in fact the cost of maintaining the current level of output and income.

Another major error involved a confusion between the cost of investment and its value.

Climate change appears to be accelerating. As it seems already a major hindrance to growth, it poses a major but unknown threat to living standards. With the benefits to growth promised by AI being equally uncertain, economic forecasts resemble guesses even more than they have historically.

This work is licensed under a Creative Commons Attribution 4.0 License.