good read -- Porn in the USA: Conservatives are biggest consumers

[David Farber <dave () farber net>]

From: Stewart Baker <stewart.baker () gmail com>
Date: Mon, Mar 2, 2009 at 9:15 AM
Subject: Re: [IP] Re: Porn in the USA: Conservatives are biggest  
consumers
To: dave () farber net


Dave,

Actually, for domestic flights, homeland security doesn't get travel
reservation data, or any other data about travelers.  The Secure
Flight program, which would allow DHS to screen travelers against the
no-fly list, has been delayed by at least five years by privacy
advocates and Congress because they were reluctant to let DHS have
access to travelers' names and birthdates.  As a result, DHS has had
to rely on the airlines to administer the no-fly list.  Since spending
money on overhead IT systems is not a high priority for the airlines,
the ultimate result is that thousands, perhaps millions, of travelers
have been subjected to no-fly delays because they have names (but not
birthdates or other data) that overlap with a terrorist's.  These
people (Ted Kennedy among them) are privacy victims, inconvenienced
daily by an irrational privacy restriction imposed by Congress and
privacy advocates.  Luckily, DHS has now finished jumping over the
hurdles erected by Congress, so this inconvenience will end in the
next year or so.

But I'm really writing to object to the underlying assumption of the
story.  It's only newsworthy if the red states are somehow revealed to
be hypocrites by the credit card data.  That's the implied conclusion
of the story: Conservatives like porn more than liberals, even though
they pretend to hate it.

But the anonymization by zip code makes that conclusion unsupportable.
 What we know is that states (or zip codes) that vote conservative
seem to buy online porn at least as much as states or zip codes that
vote liberal.  But reliably "conservative" states may be 48% liberal,
while liberal states are 52% liberal.  That's a big deal in politics,
a less big deal when looking at market stats; with a 4% swing in
politics, many market conditions can drive statistical differences
much more than political conditions.  So, here are other conclusions
that are just as plausible from the same data:

1.  Liberals actually buy more porn than conservatives, but in liberal
states, they can buy porn (for cash) by going to physical stores.  In
conservative states, those stores are banned or scarce, so liberals
have to go on line.

2.  Conservative states are less computer-savvy and thus less likely
to use file-trading networks, so people pay for porn in those states
rather than getting it free.

3.  Conservatives buy less porn but their consciences require that
they pay for it; liberals are more likely to ignore copyright
restrictions, so their consumption doesn't show up in an analysis of
credit card payments.

I don't actually know whether any of these things are so, because the
data are simply insufficient.  Of course they were insufficient for
the conclusions implied in the original press coverage of the topic,
too.  Where are all the people who usually complain when the press
commits some form of innumeracy?

Stewart



On Mon, Mar 2, 2009 at 1:51 AM, David Farber <dave () farber net> wrote:
> Begin forwarded message:
>
> From: Peter Swire <peter () peterswire net>
> Date: March 1, 2009 10:12:42 PM EST
> To: "dave () farber net" <dave () farber net>
> Subject: RE: [IP] Re:    Porn in the USA: Conservatives are biggest
> consumers
>
> Brock Meeks asks very good questions, which essentially boil down to  
> this --
> was the research done consistent with strong privacy protections?
>
> As Brock and many readers of this list know, there have been painful
> episodes where data was released for research purposes in the belief  
> that
> privacy was protected.  Then, the data was linked to an individual.   
> One
> example was when AOL released the records of online searches, and a  
> number
> of the indviduals were identified.  As a more general matter, LaTanya
> Sweeney and others have shown that "deidentified" data can often be
> "reidentified."
>
> On the other hand, research is often a Good Thing.  In the medical  
> realm,
> research can go forward consistent with the HIPAA privacy rule in  
> several
> ways -- consent by the patient, deidentification, approval by the
> Institutional Review Board, or subject to a data use agreement.
>
> Other privacy laws, and many privacy policies, do not have any  
> similar path
> to conducting research.
>
> I think policymakers, and readers of this list, should want to  
> accomplish
> the Good Thing of research and the Good Thing of privacy when  
> possible.  As
> a rough guide, the following should be considered:
>
> 1. Release data to researchers in truly deidentified form.  That may  
> have
> happened in the porn study, if the data was kept at the zip code  
> level.
>  (Then again, the list of credit card purchases may be like the AOL  
> list of
> search terms -- some purchases or search terms are idiosyncratic  
> enough
> that, combined with public records, the person can be identified.)
>
> 2. Consider creating Institutional Review Boards or the equivalent  
> so that
> good practices are followed.  It is true that IRBs can be a  
> bureaucratic
> burden.  But they also are an institutional mechanism to come up with
> protocols that meet good standards.
>
> 3. Consider releasing the data under data use agreements that bind the
> researchers.  In this way, the researcher breaches a contract if  
> names or
> other personal data are released outside the scope of the permitted
> research.
>
> Steps such as these can help us get useful research while also  
> protecting
> individuals against privacy invasions they can't control.
>
> Peter
>
> Prof. Peter P. Swire
> C. William O'Neill Professor of Law
> Moritz College of Law of the Ohio State University
> Senior Fellow, Center for American Progress
> (240) 994.4142, www.peterswire.net
>
>
> -----Original Message-----
> From: David Farber [mailto:dave () farber net]
> Sent: Sunday, March 01, 2009 6:02 PM
> To: ip
> Subject: [IP] Re: Porn in the USA: Conservatives are biggest consumers
>
>
>
> Begin forwarded message:
>
> From: "Brock N. Meeks" <bmeeks () cox net>
> Date: March 1, 2009 1:50:27 PM EST
> To: <dave () farber net>, <jwarren () well com>
> Subject: Re: [IP] Porn in the USA: Conservatives are biggest consumers
>
> Leaving the original <ahem> thrust of this message aside, my foremost
> question was "how did the researcher obtain this data?"
>
> I've known Jim Warren for a couple of decades and I know how he values
> privacy, which is why I'm curious as to why Jim didn't raise the  
> subject
> himself.
>
> In the New Scientist article, we read the researcher has a client that
> runs
> adult entertainment web sites and that this company provided the
> researcher
> "with roughly two years of credit card data from 2006 to 2008 that
> included
> a purchase date and each customer's postal code."
>
> Say what?  I'm sure the last thing these subscribers thought their
> credit
> card info would be used for was any kind of social behavior study.
>
> Perhaps the "privacy" policies of such web sites informed subscribers
> that
> such a use of their information would be possible (I've not studied  
> the
> privacy policies of such web sites, I just look at the pictures...
> Wait...)
>
> If say, oh, for example, the Airline industry, turned over this kind  
> of
> "anonymized"  credit card data to homeland security for a study of
> (whatever) I think Jim would be more than a bit concerned.
>
> Now, I suppose such a hypothetical sets up a comparison as to who is  
> the
> more trustworthy:  the adult entertainment industry, the researcher,  
> the
> airline industry or Uncle Sam.  I leave that debate to more informed
> colleagues.
>
> --Brock
>
>
>
>
>
>
> On 3/1/09 10:21 AM, "David Farber" <dave () farber net> wrote:
>
> > Begin forwarded message:
> >
> > From: Jim Warren <jwarren () well com>
> > Date: February 28, 2009 4:27:59 PM EST
> > To: Dave Farber <dave () farber net>
> > Subject: Porn in the USA: Conservatives are biggest consumers
> >
> > From the New Scientist - http://bit.ly/ZkOq
> > 27 February 2009 by Ewen Callaway
> >
> > Americans may paint themselves in increasingly bright shades of red
> > and blue, but new research finds one thing that varies little across
> > the nation: the liking for online pornography.
> >
> > A new nationwide study (pdf) of anonymised credit-card receipts  
> > from a
> > major online adult entertainment provider finds little variation in
> > consumption between states.
> >
> > "When it comes to adult entertainment, it seems people are more the
> > same than different," says Benjamin Edelman at Harvard Business
> > School.
> >
> > However, there are some trends to be seen in the data. Those states
> > that do consume the most porn tend to be more conservative and
> > religious than states with lower levels of consumption, the study
> > finds.
> >
> > ...<big snip>...
>
>
>
>
>
>
> -------------------------------------------
> Archives: https://www.listbox.com/member/archive/247/=now
> RSS Feed: https://www.listbox.com/member/archive/rss/247/
> Powered by Listbox: http://www.listbox.com
>
>
>
>
> -------------------------------------------
> Archives: https://www.listbox.com/member/archive/247/=now
> RSS Feed: https://www.listbox.com/member/archive/rss/247/
> Powered by Listbox: http://www.listbox.com



--
Stewart Baker
202-641-8670



--
Stewart Baker
202-641-8670




-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com