attitudes toward online availability of us public records
TRANSCRIPT
Sunlight or Sunburn:Attitudes toward Online Availability of US Public Records
Sean A. Munson
Daniel Avrahami, Sunny Consolvo
James Fogarty, Batya Friedman
Ian Smith
University of Michigan
Intel Labs Seattle
University of Washington
Everbread, Ltd.
public records aren’t new
• In US, campaign contribution records date
from 1910, upheld in 1976 (Buckley v Valeo).
… but this level of accessibility is.
• Long history of real estate records being
public
The Good
“Publicity is justly commended as a remedy for
social and industrial diseases. Sunlight is said to
be the best of disinfectants; electric light the most
efficient policeman” - Justice Brandeis, 1914
Cause for concern? (or at least questions)
Are these uses consistent with campaign donors’ and real estate purchasers’ expectations?
If not, people may feel that their privacy has been violated (Adams & Sasse).
Nissenbaum on public records’ increased accessibility:
“requires an examination of governing norms of appropriateness and flow to see whether and in what ways the proposed new practices measure up.”
Risks of violations
If there are violations:
• Risk of people suppressing activities in ways not
intended?
• Damage to relationships?
• Culture of mutual surveillance? (1984; Palen &
Dourish)
Aggregation of public data sets to
derive sensitive data. (Griffith & Jakobsson 2005)
Survey design
Four page survey, questions
about:
• demographics
• history and intentions for
campaign contributions and
real estate purchases
• whether they had displayed
political bumper stickers or
signs
• comfort about what of this
information was available online
• how its availability may affect
future plans
Screenshots of existing sites and
tools
Survey design
Four page survey, questions
about:
• demographics
• history and intentions for
campaign contributions and
real estate purchases
• whether they had displayed
political bumper stickers or
signs
• comfort about what of this
information was available online
• how its availability may affect
future plans
Screenshots of existing sites and
tools
Respondents
Survey sent to 1,000
households in PNW,
November 2008.
• 47 marked “return to
sender”
• 134 filled out and
returned (14% response
rate)
Somewhat older and more
liberal than general
population
Beliefs, understanding, and awareness
Though most participants offered reasons for
records being public, some could not think of a
reason
– 19% for real estate records
– 11% for campaign contributions
Differing awareness of online availability
– 73% for real estate records
– 49% for campaign contributions
Beliefs, understanding, and awareness
Real estate records: emphasis on practical
concerns (e.g., information for potential buyers,
sellers, and tax records; 60%)
Campaign contributions: emphasis on
transparency (49%)
c
Comfort with
records being
searchable
Comfort with searching by
Real estate purchases
Campaign contributions
State
City
Zip code
Neighborhood
name
Home address
Last name only
First & last name
Employer
Occupation
Wilcoxon Signed-rank test, correction for multiple tests.
Very comfortable
Not at all comfortable
c
Comfort with
records being
searchable
Comfort with searching by
Real estate purchases
Campaign contributions
State
City
Zip code
Neighborhood
name
Home address
Last name only
First & last name
Employer
Occupation
Very comfortable
Not at all comfortable
c
Comfort with
records being
searchable
Comfort with searching by
Real estate purchases
Campaign contributions
State
City
Zip code
Neighborhood
name
Home address
Last name only
First & last name
Employer
Occupation
Wilcoxon Signed-rank test, correction for multiple tests.
Comfort with
searcher
Greater comfort with searcher from
their area and a citizen than with a
searcher from outside their area or
who was a non-citizen.
Comfort if the person accessing these
details online lives
Real estate
purchases
Campaign
contributions
in the same
neighborhood as you
in the same city as you, but not in the same
neighborhood
µ = 3.21, stdev = 1.66
in the same state as
you, but not the same city
in the US and is a US citizen
outside of the US but is a US citizen
in the US, and is a US Permanent Resident
outside of the US, but is a US Permanent
Resident
in the US legally, but is not Citizen or
Permanent Resident
in the US, but not
legally and is not a Citizen or Permanent
Resident
outside of the US and is not a Citizen or
Permanent Resident
Access restrictions?
Sense from some that searchers should have
to “justify” access or pay a nominal charge to
cut down on “frivolous” access and limit it to
those who “need” to know.
… or that it should just require more effort than
visiting a website.
“city/county Admin but not where it can be
readily accessed online.”
Eliminate?
There should be right to privacy - like the secret
ballot - political affiliations & contributions
should remain private. I was major ticked off
to find my contributions, candidates' affiliations,
etc online - no one's business in the public
arena.
“there should be no access. One's political
affiliations should only be made public
voluntarily.”
Eliminate?
33% of those who wanted change said they would give less to political campaigns in the future, vs. only 3% for those who did not want access changed.
“This issue has already made us unlikely to ever donate to a presidential campaign again.”
A Dilemma
“this is a dilemma: privacy vs. protection from
influence by individuals or groups”
“I prefer privacy & I'm not sure how to address
the issue.”
A Dilemma: What to do?
Do we
• change policies?
• better educate donors and buyers?
– Why records are available
– What will be accessible (better labels?)
• reconsider what tools we build?
Privacy violations are occurring, with unknown
effects.
Future work
Update data (quite a bit has changed since 2008)
Need to measure actual behavior and effects
rather than just attitudes.
Sunlight or Sunburn:Attitudes toward Online Availability of US Public Records
Sean A. Munson
Daniel Avrahami, Sunny Consolvo
James Fogarty, Batya Friedman
Ian Smith
University of Michigan
Intel Labs Seattle
University of Washington
Everbread, Ltd.
Funded by the National Science Foundation IIS-0325035 and
Intel.
Thanks to Amanda Fondville and Charity Leang.