Action!
Chris Csikszentmihalyi
How does one transform data into social action? What combinations of
information, access, interface, interaction, and social organization
best lead to substantial social transformation? These questions are
at the heart of our research at the MIT Center for Future Civic Media,
and I had the honor of posing them to three top thinkers and
practitioners in data and social change during a plenary panel at our
conference last month.
[http://civic.mit.edu/watchlistenlearn/data-into-action-plenary-from-2010-future-of-news-and-civic-media-conference]
In challenging the speakers, I gave two models as examples. I will
call the first model the "Brockovitch"
[http://en.wikipedia.org/wiki/Erin_Brockovitch]. In it, a[n]
[investigative journalist / whistleblower / activist / regulator /
informed citizen] carefully researches and analyzes an obscure
database that was probably generated by the government to comply with
a reporting law passed years before. The investigator then uses
[statistics / data visualization] to [sue / advocate / write a story
about] [systemic discrimination / corruption / inefficiency / a sex
scandal]. This is a well known model, off-gassing heroism: the truth
is out there, follow the money, look for patterns, break it wide open.
Slap the analysis into the [evening edition / indictment / policy
brief / public domain], and use it the proof to elicit change.
The second model is more quotidian, widely distributed, and decidedly
less cinematically heroic; I will call the "Tufte"
[http://en.wikipedia.org/wiki/Edward_Tufte]. In the second model,
data that has long been available [realtime bus GPS data / credit
reports / local EPA toxics inventories] but difficult to access are
suddenly made legible through [information design / just in time
delivery / phone 'app' ]. Unlike the first model, there was no
conspiracy to uncover; rather, useful information was simply not
making its way to the citizen in a way that they could use when they
needed it.
I would like to add a third model, one that seems to be showing a lot
of promise, which I will call the "LittleSis" [http://littlesis.org/].
In this model, no suitable database exists for a particular
application. But through [aggregation / mashing up] some basic data
or categories, [informed citizens / specialists / retirees /
activists] collaboratively research and collect data, using a [wiki /
web application / edge-node database]. At that point the data will be
used similarly to the "Brokovitch," perhaps mined by journalists or
regulators. Unlike the "Brokovitch," however, users played just as
important a role in generating and contributing data as in analyzing
and leveraging it.
It should be clear from these three models that there are many
parameters involved. Who are the users? Is the data governmental and
top-down, or is it being generated by users? Once the data is
analyzed, what is the next step? Does the same entity host both the
data and the system that makes the data useful? Indeed, there are so
many parameters that it could easily be argued that these three models
represent entirely different activities, and there is no similarity
between them.
Certainly there are nuances in each of these models, nuances that
necessitate different sorts of design and implementation. But I would
argue that they have many features in common, and indeed represent
points on a spectrum. For example, if a strong site like LittleSis
were to do such a good job of identifying conspiratorial connections
and oligarchic subterfuge that lobbyists had a harder time creating
corruption in government, there would be less need for journalists to
have to muckrake through the databases at a later date. And if
real-time bus data was available as a convenience for riders waiting
in the rain, then data would automatically available to research
inequity of transportation routes based on neighborhood income. Each
of these models represents an appropriate fulcrum point at which
well-marshaled data can be leveraged to assist civil society. But
which model is likely to be the most effective in a particular
situation? If we integrate these models vertically, will that lead to
more or less impact? Are there 50 more models out there waiting to be
discovered?
While each expert on the Data Into Action panel answered the question
based on their extensive experience, all the panelists -- Ellen Miller
(Sunlight Foundation), Nick Grossman (Civic Works & OpenPlans), and
Laurel Ruma (O'Reilly and Associates) -- indicated that there is still
a lot of experimentation to be done before anyone can answer the
question with any certainty.
Then by all means, let us experiment! This list will mix social
entrepreneurs, academics, journalists, government functionaries, and
activists. Let us use this blog and this mailing list to share our
experiments and experiences, to help and challenge each other, to
build our collective knowledge, and to drive our ability to push
information technologies to build a more open, equal, and informed
society.