[geo-discuss] quasi-public goods

[Rufus Pollock]

Jo Walsh wrote:
> I dug out the 1999 paper which the OS website quotes everywhere on the
> subject of how OS data 'underpins 100 billion worth of business' (for
> which read, OS sell some data to companies which have throughput of
> 75-125 billion between them)
> http://www.ordnancesurvey.co.uk/oswebsite/aboutus/reports/oxera/
> 
> i was fascinated to see the Jamie Love connection in the citing of
> government-collected spatial information as a "quasi-public good",
> which appears in other papers that reference this one, e.g.
> http://geoinfo.uneca.org/sdiafrica/Chap_HTML/07Financing.htm
> 
> thoughts:
> - the ideas of non-excludability on which the definition of public
>   good is built don't seem to apply to digital goods. All digital
>   goods tend towards being quasi-private in the current IP climate

Public goods are defined usually as being *both* non-excludable *and* 
non-rival. Monopoly rights in information goods (patents, copyright etc) 
are expressly designed to artificially create excludability. This will 
*never* be optimal since the goods are non-rival but it is one way to 
address under-production (others being direct funding, prizes etc). 
Since, as you point out, one can change the excludability of something I 
don't always think this defn is very useful and find it better to talk 
about excludability and rivalry directly.

> - if NIMSA funded data collection is clearly separable from other geodata
>   collection, and *is* classified as a public good, then surely public
>   access to that information should also be considered a public good?

That would be misusing the term public good. Good in economist's lingo 
doesn't mean a good thing but simply a commodity (or more precisely 
things over which you can have preferences e.g. the amount of noise your 
neighbour produces on saturday night could be considered a good but 
isn't really a commodity). Geodata here is the public good. It's 
collection nor its distribution are public goods. What you can say is 
that for public goods that have been (and can be) paid for up front by 
the government it is optimal for social welfare to have access charged 
at the cost of *distribution* (note *not* the cost of collection).

> i am not an economist, and the lingo is intermittently penetrable. 
> i thought the papers might catch someone's interest though.

Thanks Jo. I'll definitely check them out.