[Wikimedia-l] Any studies on economic impact of community-produced open data?

classic Classic list List threaded Threaded
15 messages Options
Reply | Threaded
Open this post in threaded view
|

[Wikimedia-l] Any studies on economic impact of community-produced open data?

Tom Morris-5
Yesterday I was at Over the Air, the UK mobile hack day, where developers come together and try to build mobile apps and projects, commercial and open source. (I worked on something Wikimedia-related, more of which when I've made the code presentable.)

One of the things discussed was the perennial topic of opening up of UK government data. The UK government have committed to making as much data as possible openly available under open licenses (specifically the Open Government Licence, which is basically the UK government's rebadged CC BY license).

The UK government are trying to seed use of data with companies, to show an economic and social use for opening up data, and are prioritising open data releases that have some useful economic benefit. For instance, hypothetically, releasing high-quality information about public transport would allow people to develop mobile apps to help people use public transport, while releasing data about bird population studies might be of less commercial importance. The government seem to be leaning towards putting out the commercially useful material first.

One thing that came up in discussion was whether or not anyone has ever done any economic impact studies on Wikipedia and other community-produced open data and open content projects (OpenStreetMap, other Wikimedia projects etc.)? If civic society groups like Wikimedians, OpenStreetMappers, MySociety.org people etc. want to convince governments to put more data out, it'd be helpful to show the economic effects of this, or to have people who are trying to convince government to put the data out to have access to this kind of information so they can make better decisions (cue cynicism here).

There are such reports on the economic impact of open data releases by governments[1]. But I was wondering if we'd seen anything for community-produced data. I know Apple are now using OpenStreetMap within iPhoto[2] and have long had Wikipedia as part of OS X's Dictionary.app. The Foursquare website has just switched over to OSM (the iOS and Android apps both still use Google Maps).

Obviously, the success of Wikipedia has affected the previously dominant players in the encyclopaedia market: Britannica, Encarta, World Book etc. But it is also providing all sorts of much harder to see effects by reducing costs for businesses and organisations. The BBC reuse Wikipedia content within their music and wildlife websites, thus reducing the amount they have to pay for content (or, more charitably, enabling them to do projects that they wouldn't otherwise be able to do). Wikipedia makes Google more useful, and Google have often said "anything that makes the web better makes Google better". Might a decent Wikipedia in a small language effect the market in that language community for technology? If people can actually read stuff in language X, does that increase the demand for computers/mobile phones/Internet access in language X speakers?

Has there been any studies of these kinds of economic impact of Wikimedia projects and other open content/open data projects?

I have seen some discussion of this in the OSM community.[3] Unfortunately, Google is failing me on searching for material on the economic impact of Wikipedia: I just get lots of Wikipedia articles with titles of the form "Economic impact of X". I also couldn't find anything on the Research Index section of Meta.

My eventual interest in this is whether or not there are potential ways governments could work with projects like Wikipedia, chapters like WMUK and with individual volunteers as part of their open data strategy: they seem to want to do likewise with commercial organisations because of the obvious economic benefits that some of that data has. There's potential for a kind of three-way thing, with governments working with both a commercial partner and with a community partner (like a Wikimedia chapter) to produce symbiotically beneficial data.[4]

[1] see http://wiki.linkedgov.org/index.php/The_economic_impact_of_open_data
[2] http://www.theverge.com/2012/5/4/2998428/apple-iphoto-ios-openstreetmap-credit
[3] http://lists.openstreetmap.org/pipermail/talk/2010-September/053947.html
[4] I hate myself when I write sentences like that.

--
Tom Morris
<http://tommorris.org/>



_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Chris Keating-2
>
> One thing that came up in discussion was whether or not anyone has ever
> done any economic impact studies on Wikipedia and other community-produced
> open data and open content projects (OpenStreetMap, other Wikimedia
> projects etc.)? If civic society groups like Wikimedians,
> OpenStreetMappers, MySociety.org people etc. want to convince governments
> to put more data out, it'd be helpful to show the economic effects of this,
> or to have people who are trying to convince government to put the data out
> to have access to this kind of information so they can make better
> decisions (cue cynicism here).
>

There is a basic problem here in that it's difficult to account for
non-traded goods. Since no-one pays for Wikipedia, the value to the economy
appears to be zero.

An alternative methodology would be to account for the value that would be
required to replace Wikipedia if it didn't exist. As an example of this
methodology you could take the traded price of a Wikipedia substitute (e.g.
Britannica Online is £50 a year) and multiply that by the number of users,
which I'd estimate at 30 million in the UK. So the hidden value to the UK
economy of Wikipedia could be as high as £1.5 billion every year....

But this might not be the most helpful answer to you. :-)

Chris
_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Thomas Dalton
On 3 June 2012 13:52, Chris Keating <[hidden email]> wrote:
> An alternative methodology would be to account for the value that would be
> required to replace Wikipedia if it didn't exist. As an example of this
> methodology you could take the traded price of a Wikipedia substitute (e.g.
> Britannica Online is £50 a year) and multiply that by the number of users,
> which I'd estimate at 30 million in the UK. So the hidden value to the UK
> economy of Wikipedia could be as high as £1.5 billion every year....

The problem with that kind of approach is that you are equating price
and value. When a sale takes place, it happens at a price somewhere
between the value to the buyer and the value to the seller (although
the value to the seller is a little difficult to define for something
like an online subscription where the unit cost is essentially zero).
That means the value of a Britannica subscription for those that buy
one is actually more than £50 (otherwise they wouldn't have bought it
- they would have been at least as happy just keeping the £50).
However, for those that don't buy one (and, even if Wikipedia
vanished, most of our readers wouldn't buy one), the value is less
than £50 (that's why they don't buy it).

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

erikzachte
I doubt most subscribers to Britannica Online access it almost every day, at home and at work, for even most mundane information needs.
Wikipedia is much more filled with practical information that helps us save time and avoid costly mistakes every day.
Is Britannica Online comparable in content to the paper edition? Ask any owner of that paper edition how often they grabbed a volume from the shelf, even before Wikipedia came along.
Mainly for this reason I doubt people would mass subscribe to Britannica when Wikipedia disappeared.

Given that Google and other search engines serve our pages often in the top 5 results, we might perhaps derive our economic value from their gross profit (would 5% be a conservative enough wild guess)?
Google's gross profit is about $26 billion per year [1]. 5% added value would equal $1.3 billion per year and that is only the Google part, and they are involved in 45% of our external requests [2].

(still this leaves Thomas argument that money spent does not equal financial gain)

At first Wikimania, at Frankfurt, Jimmy gave an estimate how much money we could have earned from advertisements, and it was already more than a million per month in 2005.

[1] http://www.wolframalpha.com/input/?i=gross+profit+google
[2] http://stats.wikimedia.org/wikimedia/squids/SquidReportGoogle.htm

Erik Zachte

-----Original Message-----
From: [hidden email] [mailto:[hidden email]] On Behalf Of Thomas Dalton
Sent: Sunday, June 03, 2012 3:08 PM
To: Wikimedia Mailing List
Subject: Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

On 3 June 2012 13:52, Chris Keating <[hidden email]> wrote:
> An alternative methodology would be to account for the value that
> would be required to replace Wikipedia if it didn't exist. As an
> example of this methodology you could take the traded price of a Wikipedia substitute (e.g.
> Britannica Online is £50 a year) and multiply that by the number of
> users, which I'd estimate at 30 million in the UK. So the hidden value
> to the UK economy of Wikipedia could be as high as £1.5 billion every year....

The problem with that kind of approach is that you are equating price and value. When a sale takes place, it happens at a price somewhere between the value to the buyer and the value to the seller (although the value to the seller is a little difficult to define for something like an online subscription where the unit cost is essentially zero).
That means the value of a Britannica subscription for those that buy one is actually more than £50 (otherwise they wouldn't have bought it
- they would have been at least as happy just keeping the £50).
However, for those that don't buy one (and, even if Wikipedia vanished, most of our readers wouldn't buy one), the value is less than £50 (that's why they don't buy it).

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l



_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Fred Bauder-2
We are serving up our product at cost. That we might, by rent seeking,
charge a much higher price is due more to the geometry of possible rent
seeking opportunities than to economic factors. Rent seeking, or profits,
is an optional cost imposed by political decision.

This situation can exist because a tiny fraction of the global workforce
will work simply for a sense of accomplishment, or whatever. It has no
applicability to the vast majority of workers who will toil only for
reasonable compensation.

Fred

> I doubt most subscribers to Britannica Online access it almost every day,
> at home and at work, for even most mundane information needs.
> Wikipedia is much more filled with practical information that helps us
> save time and avoid costly mistakes every day.
> Is Britannica Online comparable in content to the paper edition? Ask any
> owner of that paper edition how often they grabbed a volume from the
> shelf, even before Wikipedia came along.
> Mainly for this reason I doubt people would mass subscribe to Britannica
> when Wikipedia disappeared.
>
> Given that Google and other search engines serve our pages often in the
> top 5 results, we might perhaps derive our economic value from their
> gross profit (would 5% be a conservative enough wild guess)?
> Google's gross profit is about $26 billion per year [1]. 5% added value
> would equal $1.3 billion per year and that is only the Google part, and
> they are involved in 45% of our external requests [2].
>
> (still this leaves Thomas argument that money spent does not equal
> financial gain)
>
> At first Wikimania, at Frankfurt, Jimmy gave an estimate how much money
> we could have earned from advertisements, and it was already more than a
> million per month in 2005.
>
> [1] http://www.wolframalpha.com/input/?i=gross+profit+google
> [2] http://stats.wikimedia.org/wikimedia/squids/SquidReportGoogle.htm
>
> Erik Zachte
>
> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of Thomas
> Dalton
> Sent: Sunday, June 03, 2012 3:08 PM
> To: Wikimedia Mailing List
> Subject: Re: [Wikimedia-l] Any studies on economic impact of
> community-produced open data?
>
> On 3 June 2012 13:52, Chris Keating <[hidden email]> wrote:
>> An alternative methodology would be to account for the value that
>> would be required to replace Wikipedia if it didn't exist. As an
>> example of this methodology you could take the traded price of a
>> Wikipedia substitute (e.g.
>> Britannica Online is £50 a year) and multiply that by the number of
>> users, which I'd estimate at 30 million in the UK. So the hidden value
>> to the UK economy of Wikipedia could be as high as £1.5 billion every
>> year....
>
> The problem with that kind of approach is that you are equating price and
> value. When a sale takes place, it happens at a price somewhere between
> the value to the buyer and the value to the seller (although the value to
> the seller is a little difficult to define for something like an online
> subscription where the unit cost is essentially zero).
> That means the value of a Britannica subscription for those that buy one
> is actually more than £50 (otherwise they wouldn't have bought it
> - they would have been at least as happy just keeping the £50).
> However, for those that don't buy one (and, even if Wikipedia vanished,
> most of our readers wouldn't buy one), the value is less than £50
> (that's why they don't buy it).
>
> _______________________________________________
> Wikimedia-l mailing list
> [hidden email]
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
>
>
>
> _______________________________________________
> Wikimedia-l mailing list
> [hidden email]
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
>



_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Thomas Dalton
On 3 June 2012 18:25, Fred Bauder <[hidden email]> wrote:
> We are serving up our product at cost. That we might, by rent seeking,
> charge a much higher price is due more to the geometry of possible rent
> seeking opportunities than to economic factors. Rent seeking, or profits,
> is an optional cost imposed by political decision.

Yes, but we're not talking about cost. We're talking about value. Just
because we are able to produce Wikipedia at very little cost doesn't
mean it has very little value.

> This situation can exist because a tiny fraction of the global workforce
> will work simply for a sense of accomplishment, or whatever. It has no
> applicability to the vast majority of workers who will toil only for
> reasonable compensation.

I think most people have an altruistic streak - looking after your
parents in their retirement is altruistic, for example. We're a little
unusual in that we're happy to work for the benefit of humanity
generally, rather than some specific, tangible beneficiaries, but
altruism in general isn't unusual.

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
???
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

???
In reply to this post by erikzachte
On 03/06/2012 18:10, Erik Zachte wrote:

> I doubt most subscribers to Britannica Online access it almost every
> day, at home and at work, for even most mundane information needs.
> Wikipedia is much more filled with practical information that helps
> us save time and avoid costly mistakes every day. Is Britannica
> Online comparable in content to the paper edition? Ask any owner of
> that paper edition how often they grabbed a volume from the shelf,
> even before Wikipedia came along. Mainly for this reason I doubt
> people would mass subscribe to Britannica when Wikipedia
> disappeared.
>

Personally I used the paper version whenever I wanted to check things
out. That probably wasn't as often as I use the internet to check stuff
because I didn't have that much need then. I still use the CD version of
Britannica if I want to be reasonably sure of some idea before
researching it further.


> (still this leaves Thomas argument that money spent does not equal
> financial gain)
>


There is little value to the economy from being able to settle a pub
bet. Or to furnish an argument on USENET or facebook.

One would, I think need to establish that the work was being used for
educational purposes that 'matter' educationally, and those that had
read the material were better as a result.

Unfortunately the coverage of history, philosophy, geography, and arts
subjects are unreadable and almost worthless.

EXAMPLE:

      Charles IV died in 1328, leaving only a daughter,
      and an unborn infant who would prove to be a girl.

no modern writer writes like that.

EXAMPLE:

     The senior line of the Capetian dynasty thus ended,
     creating a crisis over the French succession.

There was no crises in France about it. Philip VI took over in the same
way that Charles IV had succeeded. Both examples come from two
successive sentences in the 100 years war article. The rest of the
article is riddled with such stupidities.


> At first Wikimania, at Frankfurt, Jimmy gave an estimate how much
> money we could have earned from advertisements, and it was already
> more than a million per month in 2005.
>



To Government that would be a net loss as they would otherwise be able
to collect taxes on profits.



_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

David Gerard-2
On 4 June 2012 11:23, ??? <[hidden email]> wrote:

> One would, I think need to establish that the work was being used for
> educational purposes that 'matter' educationally, and those that had read
> the material were better as a result.


AIUI, weekdays office hours are our peak access period, and Wikipedia
generally isn't blocked in offices the way Facebook, etc. often are.
This suggests it's good for *something* economically.


- d.

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Thomas Dalton
On 4 June 2012 13:57, David Gerard <[hidden email]> wrote:
> AIUI, weekdays office hours are our peak access period, and Wikipedia
> generally isn't blocked in offices the way Facebook, etc. often are.
> This suggests it's good for *something* economically.

It's good for lowering the productivity of offices! I occasionally
look things up on Wikipedia at work that are actually about my work,
but usually it's to settle a debate that has nothing at all to do with
work.

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
???
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

???
In reply to this post by David Gerard-2
On 04/06/2012 13:57, David Gerard wrote:

> On 4 June 2012 11:23, ???<[hidden email]>  wrote:
>
>> One would, I think need to establish that the work was being used for
>> educational purposes that 'matter' educationally, and those that had read
>> the material were better as a result.
>
>
> AIUI, weekdays office hours are our peak access period, and Wikipedia
> generally isn't blocked in offices the way Facebook, etc. often are.
> This suggests it's good for *something* economically.
>
>

With access across 24 time zones how do you pick out weekday office
hours, as opposed to evening access?


_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Andre Engels
In reply to this post by Thomas Dalton
On Mon, Jun 4, 2012 at 3:17 PM, Thomas Dalton <[hidden email]> wrote:
> On 4 June 2012 13:57, David Gerard <[hidden email]> wrote:
>> AIUI, weekdays office hours are our peak access period, and Wikipedia
>> generally isn't blocked in offices the way Facebook, etc. often are.
>> This suggests it's good for *something* economically.
>
> It's good for lowering the productivity of offices! I occasionally
> look things up on Wikipedia at work that are actually about my work,
> but usually it's to settle a debate that has nothing at all to do with
> work.

Even then it could be good for productivity, by decreasing the amount
of time you and your colleagues spend over looking up who is right in
their debate (although probably that gained time will be used for more
debates, so the net effect on productivity will be close to zero).

--
André Engels, [hidden email]

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Andre Engels
In reply to this post by Tom Morris-5
I'll take another route (although probably just as meaningless as most
others). The normal way of generating money over the net is through
advertisements. How much would Wikipedia make in advertisements, would
they use them? Using a conservative price of 0,5 cents per pageview,
and using the data from
http://stats.wikimedia.org/wikimedia/squids/SquidReportRequests.htm I
get that just the value of having Wikipedia for one month equates 1
billion dollars (American style billions).


--
André Engels, [hidden email]

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Andrew Gray-3
In reply to this post by ???
On 4 June 2012 18:00, ??? <[hidden email]> wrote:

> With access across 24 time zones how do you pick out weekday office hours,
> as opposed to evening access?

Using non-English projects would give a more clear result here, I
think - speakers of, say, Italian or German are more concentrated in a
narrow time-zone band than speakers of English. To a first
approximation, 90-95% of the access by from speakers of those
languages is likely to take place in CET; you might well get the same
results for, say, Japanese.

Using Portuguese or Spanish would give you much less clear results,
comparable to English.

I don't know if anyone has graphed this, but I'd like to see the results :-)

--
- Andrew Gray
  [hidden email]

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

Erik Zachte-3
I did graph this time-zone dependency for my Wikimania 2005 talk:
http://tinyurl.com/cz9x4cn
See section "What wikipedians tell us when they are sleeping"

The second line plot shows clearly how two time zones contribute to Spanish
and Portuguese and their relative weight.

Erik Zachte


-----Original Message-----
From: [hidden email]
[mailto:[hidden email]] On Behalf Of Andrew Gray
Sent: Tuesday, June 05, 2012 12:15 AM
To: Wikimedia Mailing List
Subject: Re: [Wikimedia-l] Any studies on economic impact of
community-produced open data?

On 4 June 2012 18:00, ??? <[hidden email]> wrote:

> With access across 24 time zones how do you pick out weekday office
> hours, as opposed to evening access?

Using non-English projects would give a more clear result here, I think -
speakers of, say, Italian or German are more concentrated in a narrow
time-zone band than speakers of English. To a first approximation, 90-95% of
the access by from speakers of those languages is likely to take place in
CET; you might well get the same results for, say, Japanese.

Using Portuguese or Spanish would give you much less clear results,
comparable to English.

I don't know if anyone has graphed this, but I'd like to see the results :-)

--
- Andrew Gray
  [hidden email]

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] Any studies on economic impact of community-produced open data?

David Richfield
In reply to this post by Thomas Dalton
On Mon, Jun 4, 2012 at 3:17 PM, Thomas Dalton <[hidden email]> wrote:
> On 4 June 2012 13:57, David Gerard <[hidden email]> wrote:
>> AIUI, weekdays office hours are our peak access period, and Wikipedia
>> generally isn't blocked in offices the way Facebook, etc. often are.
>> This suggests it's good for *something* economically.
>
> It's good for lowering the productivity of offices! I occasionally
> look things up on Wikipedia at work that are actually about my work,
> but usually it's to settle a debate that has nothing at all to do with
> work.

Maybe that's true for you, but I often look up work-relevant stuff in
Wikipedia.  Where it's important, I can follow references and look up
other sources.

Also, think about the way doctors use it: it's a very useful
aide-de-memoire for most doctors when it comes to stuff they don't use
on a daily basis.  They have the knowledge and skills to determine
whether the information is correct, and don't rely on it for critical
decisions, but if a doctor needs a remindor of which receptor is
affected by the autoimmune system in myasthenia gravis,
http://en.wikipedia.org/wiki/Myasthenia_gravis#Pathophysiology will
help.

--
David Richfield
[[:en:User:Slashme]]
+27718539985

_______________________________________________
Wikimedia-l mailing list
[hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l