Sometimes "the hunt" for any ideological differences among economists, seems as though a time honored tradition! In recent months, Paul Romer has also jumped into the fray. In particular, his assumptions regarding Nick Rowe's attitude towards rival goods are still difficult for me to comprehend.
The definitions of rival and non rival goods are hardly "set in stone", plus these designations apply to multiple aspects of economic thought. It's good that rival and non rival resource capacity is being discussed, because this subject definitely deserves a closer look. Not only do rival and non rival designations establish patterns for resource use in the short run, the designations hold important clues as to inevitable shifts in those patterns over time. (For any new readers, I wrote several posts on rival and non rival time value in July, here, here and here.)
One way to think about a rival designation for a given resource, is its potential capacity to generate sets of expectations for patterns of economic flows. Reliable economic flows are of course desirable when possible - if not always a rational expectation on the part of the designated resource. When do rival resource designations generate positive resource flows, in aggregate? It depends, and most important is the fact that a rival designation may not necessarily guarantee a permanent (long term reliable) resource flow.
Worse, arbitrary designations can occasionally usurp monetary flows from other resource structures, especially when designations are intended to create artificial scarcity. So long as a given equilibrium pattern continues to expand through traditional production, these arbitrary designations can be smoothed over through other resource capacity. But eventually, when large percentages of private property are preset for equilibrium "high bets" on resource flows, more resources end up underutilized - or in some instances scarcely utilized at all - if and when given bets prove too large.
Time value aggregates are the least utilized resource (in terms of what matters), in present day general equilibrium. As skills value demanded more resources from the marketplace, time value gradually became subservient to skills value. However, time value needs valid connections to monetary capacity, in order for disparate groups to maintain trust among one another. Time value is naturally rival in the sense of place, time and specific circumstance for specific knowledge use. Skills value is also treated as rival. But when skills value removes too much of the marketplace for time value, coordination patterns are disturbed in ways which become increasingly difficult for monetary policy - and governments - to remedy.
Knowledge use is naturally non rival, because it has the potential to freely spread and also be adaptable to different circumstance. However, when skills value is compensated instead of time value, knowledge use tends to be utilized primarily in rival settings. The main problem in this regard is that monetary compensation cannot continue indefinitely for skills value in asymmetric patterns without causing problems for both monetary policy and knowledge use.
Any smoothly functioning economic environment will have elements of both rival and non rival structure, in order for resource flows to be maintained. In her lifetime, Elinor Ostrom was able to highlight local coordination efforts in particular for tradable goods which were effectively managed by local groups. The challenge for the present is to extend similar coordination efforts into the realm of knowledge based services which take place in a time based framework.