Skip to content
Last Dictionary
Plane
Society, Ethics & EconomyMachine Learning

demographic parity

nounAlso: statistical parity, group parity, independence (fairness criterion)

In plain wordsA fairness test that asks whether a system says yes (approves the loan, shortlists the applicant) at the same rate for every group, whatever other differences exist between them.

  1. A group fairnessgroup fairness, n.: Measuring fairness by comparing how a system treats whole groups, such as women and men, on average: their approval rates, their error rates or what their scores mean. criterion requiring a classifier's positive-decision rate to be equal across groups defined by a protected attributeprotected attribute, n.: A personal trait, such as race, sex, religion, age or disability, that the law says must not be used to treat people unfairly. A; equivalently, the prediction is statistically independent of A. In practice it is relaxed to a maximum difference or a minimum ratio, as in the four-fifths rule used as evidence of disparate impactdisparate impact, n.: When a rule or system that treats everyone alike on paper still lands much harder on one group, such as women or older workers; the law can forbid this even if no one intended it.. Demographic parity ignores the true outcome: if base rates genuinely differ between groups, even a perfect classifier violates it, and it can be met by a classifier that accepts qualified members of one group but arbitrary members of another. Its defenders favor it where the outcome labels themselves reflect past discrimination.

    Demographic parity would require the hiring screen to advance the same share of women and men.

P(Y^=1∣A=a)=P(Y^=1∣A=b)P(\hat{Y}=1 \mid A=a) = P(\hat{Y}=1 \mid A=b)

Origin

The criterion was formalized in discrimination-aware data mining by Toon Calders, Faisal Kamiran, Mykola Pechenizkiy and Sicco Verwer (2009–2010), and analyzed as 'statistical parity' by Cynthia Dwork and colleagues (2011–2012), who showed its weaknesses. The name 'demographic parity' was established by October 2016, when Moritz Hardt, Eric Price and Nathan Srebro called it a common conception of non-discrimination, and it became the usual machine-learning label; its first use has not been traced.

Related

See also
  • group fairnessgroup fairness, n.: Measuring fairness by comparing how a system treats whole groups, such as women and men, on average: their approval rates, their error rates or what their scores mean.
  • equalized oddsequalized odds, n.: A fairness test that asks whether a system makes mistakes at the same rates for every group: wrongly flagging innocent people, and wrongly missing real cases, equally often.
  • equal opportunityequal opportunity, n.: A fairness test that asks whether people who truly deserve a yes (who would repay the loan or succeed in the job) get one at the same rate in every group.
  • predictive paritypredictive parity, n.: A fairness test that asks whether a system's verdict means the same thing for every group: of the people it flags as high risk, the same share really are, whatever their group.
  • disparate impactdisparate impact, n.: When a rule or system that treats everyone alike on paper still lands much harder on one group, such as women or older workers; the law can forbid this even if no one intended it.
  • fairness impossibility theoremfairness impossibility theorem, n.: A mathematical result showing that when groups differ in how often an outcome really occurs, no imperfect prediction system can be fair by every common measure at once.
Contrast with
  • individual fairnessindividual fairness, n.: The idea that a fair system should treat similar people similarly: two applicants who are alike in every way that matters should get similar outcomes.

Sources

  1. Calders, T. & Verwer, S. (2010). Three naive Bayes approaches for discrimination-free classification. Data Mining and Knowledge Discovery 21(2), 277–292.
  2. Dwork, C., Hardt, M., Pitassi, T., Reingold, O. & Zemel, R. (2012). Fairness Through Awareness. Proc. ITCS 2012. arxiv.org/abs/1104.3913
  3. Hardt, M., Price, E. & Srebro, N. (2016). Equality of Opportunity in Supervised Learning. Advances in Neural Information Processing Systems 29. arxiv.org/abs/1610.02413

Mentioned in

algorithmic fairnessalgorithmic fairness, n.: The effort to make automated decisions fair: to say in numbers what fair treatment means, measure whether a system achieves it, and correct the system if it does not., disparate impactdisparate impact, n.: When a rule or system that treats everyone alike on paper still lands much harder on one group, such as women or older workers; the law can forbid this even if no one intended it., fairness impossibility theoremfairness impossibility theorem, n.: A mathematical result showing that when groups differ in how often an outcome really occurs, no imperfect prediction system can be fair by every common measure at once., protected attributeprotected attribute, n.: A personal trait, such as race, sex, religion, age or disability, that the law says must not be used to treat people unfairly., counterfactual fairnesscounterfactual fairness, n.: A fairness test that asks: had this person been of another race or sex, with everything that would have followed from that, would the system have decided the same way?, Cynthia DworkCynthia Dwork, prop. n.: An American computer scientist who co-invented differential privacy, the leading mathematical definition of privacy in data analysis, and helped found the study of fairness in algorithms., equal opportunityequal opportunity, n.: A fairness test that asks whether people who truly deserve a yes (who would repay the loan or succeed in the job) get one at the same rate in every group., equalized oddsequalized odds, n.: A fairness test that asks whether a system makes mistakes at the same rates for every group: wrongly flagging innocent people, and wrongly missing real cases, equally often., group fairnessgroup fairness, n.: Measuring fairness by comparing how a system treats whole groups, such as women and men, on average: their approval rates, their error rates or what their scores mean., individual fairnessindividual fairness, n.: The idea that a fair system should treat similar people similarly: two applicants who are alike in every way that matters should get similar outcomes., predictive paritypredictive parity, n.: A fairness test that asks whether a system's verdict means the same thing for every group: of the people it flags as high risk, the same share really are, whatever their group.