2017 Measuring Compliance 679 That nondiscrimination is a fundamental human rights principle has not gone unacknowledged. Indeed, it could potentially affect the substantive interpretation as to whether the duty-bearer is compliant or in violation to a significant extent. Recall the scale of 0 to 1 introduced in Section III. When the test reveals full attainment of the welfare basket is impossible within the given resource set (i.e. is <1), it is not sufficient to draw full conclusions with respect to compliance or violation from looking at the equivalence, or difference, between the expected and actual attainment levels only. Rather, the duty-bearer has an additional obligation to justify that its reasoning for prioritizing the proportion of its community for whom the basket has been attained over the proportion for whom it has not is adequate and reasonable. The question of whether a duty-bearer has complied with its obligations under impossibility conditions must, therefore, be examined in light of the means through which the result has been produced. For example, who are the excluded? Are they women, children, or people of a certain ethnicity? Are the individual elements within the welfare basket represented fairly or are there some goods, services, and/or facilities that are over-represented while others are completely ignored? If a further assessment reveals that the welfare basket has been provided according to sound criteria, the signal of compliance holds. If, on the other hand, the duty-bearer’s justification for denying access to the welfare basket is deemed unreasonable, then the signal of compliance folds. Discrimination in social outcomes could be incorporated into the econometric model if it could be measured systematically over space and time. However, since discrimination tends to manifest along multiple lines (class, race, gender, for example), it is not amenable to capture by a one-dimensional measure. And even if it were amenable to capture, it is unlikely that sufficiently disaggregated data be available to capture it with any validity or reliability. Obviously, the signal of compliance or violation would need further in-country analysis to assess whether it is in fact a good measure of social rights fulfillment. But, for now at least, this signal and the methodology through which it is produced is a clear move forward over existing approaches that attempt to deal with the maximum available resources dilemma. Economics certainly has something to offer.