I like the paper by Kuben Naidoo in the attached CABRI volume. It helps put capital investment in perspective.
Conventionally, we often think of a combination of recurrent plus capital spending as leading to the delivery of services, but this is not entirely accurate: it is the stock of public assets plus recurrent spending that deliver services, while capital spending is concerned with adding increments to this stock. Looked at in these terms, the cost of delivering public services may not be seen as expenditure on public investment (a stock concept) but rather as the cost of using up a part of the stock of fixed capital over the budgeting period (a flow concept) . This is the thinking underlying public sector accrual budgeting and accounting, which mimic the approach used in the private sector. GFS 2001 adopts this approach, so that ‘consumption of fixed capital’ is included in the economic classification of expenses, not capital expenditure (which appears as changes in the public sector balance sheet – acquisition of non-financial assets).
I'm not recommending universal adoption of accrual accounting - this should probably be fairly low on the list of priorities for most countries - but it is nevertheless useful to conceptualise things in these terms, not the least because of the implied emphasis on the mundane business of maintaining and renewing the existing capital stock, before investing in exciting new projects.
