The study of software economics is not yet mature. For many years, the lines of code (LOG) metrics has tended to conceal major software cost drivers such as the production of requirements, plans, specifications, manuals, and other paper documents. The advent of function point metrics in the late 1970s allowed us to explore the measurement of such noncoding activities, none of which could be properly explored or normalized using LOC metrics. Indeed, we now know that on some projects (such as large defense systems) the cost to produce paper documents is twice as much as the cost to produce the code itself. The ability to measure all activities associated with software production has led to the concept of activity-based studies of software cost. The paper discusses the approach.