Why Economists Often Under-Value Privacy

advertisement
Why Economists Often
Under-Value Privacy
Peter Swire
Federal Trade Commission Staff Briefing
May 28, 2015
Overview of the Talk
 I wrote in 2013 about economists: “My appreciation of this different
view arose when I was in government, discussing privacy with
people from many intellectual and political backgrounds. In these
conversations, I came to believe that there was one important
predictor of people who did not ‘get’ the privacy issues. That
predictor was having received graduate training in economics.”
 Some recent policy discussions have suggested that economic
analysis, preferably quantifiable, should be sole or dominant way to
assess privacy issues
 Today:
 Why economists don’t value privacy as much as others
 Some possible answers from within economics
 Some answers from outside of economics
My Critique is Founded in My Own
Economics Experience
 My view that economists generally under-value privacy was
developed within my own extensive experience in economics
 Economics major in college, plus grad work
 Lots of my scholarship is law and economics
 Special Assistant to President Obama for Economic Policy,
2009-2010 (Larry Summers)
 Earlier versions of today’s talk:
 “Efficient Confidentiality for Privacy, Security, and Confidential
Business Information” (Brookings 2003)
 “Privacy and the Use of Cost/Benefit Analysis”, FTC Workshop
2003
 I gave a version of this talk at Penn Law School this spring
Privacy and Neo-Classical Economics
 The model of the competitive market assumes perfect information
 Limits on information to match buyers and sellers make the
market less competitive, reducing efficiency
 Efficiency is the primary goal
 Posner and other economists thus see privacy limits as
presumptively inefficient & contrary to the single most important
model in economics
 For an economist, some special justification is needed to
overcome the presumption of inefficiency of data limits
Economics Alternatives: Behavioral &
Experimental Economics
 Acquisti and others have challenged the neo-classical approach in a
variety of ways
 Various biases and heuristics are used by individuals, so more
information does not necessarily help achieve an efficient outcome
 How choices are framed (endowment effect, etc.) can have large
effect on outcome, and online businesses generally control the
frame
 In my view, these are promising empirical and theoretical critiques of
the neo-classical approach
 Acquisti: “it is not possible to conclude unambiguously whether
privacy protection entails a net ‘positive’ or ‘negative” change in
purely economic terms: its impact is context specific.”
 Many economists, though, fundamentally start with the model of
perfect competition having perfect information - in a complex world,
it’s a simpler model to keep in mind
An Economics Counter-Narrative of
Property (1)
 Many economists, interestingly, favor copyright laws and other
intellectual property
 Copying is a limit on information transfer, but property rights are
protected by copyright law against “theft”
 Similar rationale for the combination to a safe (or SSN or other
data that is a “leak” and can harm me) – inefficient to let the
criminal steal my property – privacy as an intermediate good
(Joe Farrell), where don’t want data flow because it facilitates
identify or other theft
An Economics Counter-Narrative of
Property (2)
 A key point is the baseline
 Coasian analysis, with high negotiating costs, key is who has the
property right
 If assume individual owns, then large costs from privacy
invasions; HIPAA Privacy Rule CBA makes this point.
Endowment effect heightens this effect.
 If assume business owns, then large costs from restrictions on
flow
 We have high transaction costs, so allocation of the property
right matters: opt-in vs. opt-out matters
 Economics, in a Coasian view, does not have much to say about
to whom the allocation should go initially
 Interestingly, this insight makes the European approach easier
for an economist to understand. The (property/fundamental)
right is assigned to the individual.
Critiques from Outside of Economics –
the Role of “Power” (1)
 Much writing about privacy addresses possible mis-uses of power
 Government/citizen
 Employer/employee
 Holder of dossier/subject of surveillance
 My experience – if you ask people what they are concerned
about, the answer is often “they will know things about me and
have power over me”
 Economists, by contrast, have a very narrow definition of power
 “Market power” where monopoly exists
 Chicago School of antitrust rarely finds that
Economists and Power (2)
 Economists see horizontal, contractual relationships where others
see vertical, power relationships. U.S. has reacted in other settings
by putting limits on freedom of contract.
 Corporation as a “nexus of contracts” among consenting parties,
Jensen & Meckling
 Others see power relationship of top managers vs. mere
employees
 Employees should have “freedom of contract” for labor
 Others support minimum wage, maximum hour laws, and
could have minimum standards for privacy
 Others think it was not voluntary “assumption of the risk”
when a factory worker had a limb smashed by an unsafe
assembly line
 Consumers voluntarily accept EULAs, T&Cs
 Others see adhesion contracts that merit regulatory scrutiny;
in part due to power imbalance, have the whole realm of
consumer protection law
Other Things Economist Don’t Address
That Others Do
 Anita Allen recent list included: civility, human dignity, limited
government, toleration, autonomy, individualism; self-expression,
reputation, repose, intellectual life, intimacy and formality Plus,
possible long-run effects of pervasive surveillance
 Revised Sunstein version of CBA recognizes importance of being
systematic, but go well beyond quantitative:
 Short run vs. long run, and the importance of avoiding a long-run
slide into a surveillance society. With discounting, may not get
large numbers for large harms in the future.
 Privacy worries are hard to quantify, and literature on “dwarfing
of soft variables” shows that quantifiable items over-valued
 Rights don’t count: if privacy is a human right, violations of those
rights are often not included in the cost/benefit analysis
Conclusion
 Economists are trained to be skeptical of limits on information flows,
because optimal market has perfect information
 Economists may also consult their own intuitions and believe others
have the same view – interesting to see empirical research about
privacy attitudes of economists vs. others
 Internal critiques – multiple critiques from within economics show
reason for caution in applying the “perfect information” intuition
 External critiques – non-economists can say: “There are more things
in heaven and earth, Horatio, than are dreamt in your philosophy.”
 My suggestion – use the economic techniques that have served the
FTC well, but do so with humility
 Avoid believing that an economics-only approach, or a quantitative
approach, is the optimal regulatory strategy
Download