Overview

Abstract

The second issue of *Foundations* carries eight entries in two movements. The first movement takes up the statistical concepts that the inaugural issue used without naming, beginning from the loose end it left: a hedonic regression that missed one of eight Pacifica comparables by $46,455. Four entries form a chain: the first names the target, and the other three are different kinds of failure. Conditional Expectation names what a regression estimates and establishes the identity on which the movement turns. Omitted Variable Bias gives the formal account of what happened to issue 1's coefficients, the case where needed information is present but unused. Endogeneity generalizes it, showing that measuring an omitted variable cures the omission without necessarily curing the correlation between regressor and structural error. Parameter Identification asks whether a rate could be recovered at all, exhibiting a sample where including the same variable fixes nothing because the information is absent rather than unused. The second movement is legal-conceptual, forced by this issue's survey of recent court decisions. The Scheme, and Effects on Value is the hinge between the movements: it takes the project for which property is compulsorily acquired as the reference for attributing value effects, and states in the first movement's terms that the law fixes the form of the valuation counterfactual, a tribunal fixes its extent, and the identification of no-scheme value from scheme-world data stays with the appraiser. The Purposes of Valuation sets out the journal's classification of practice, eight purposes, beneath each of which the property interest, the governing law and the basis of value are settled together, in no fixed order, with methodology beneath those, with the quarter's verified instruments placed under their purposes, the purposes without one reported empty rather than filled, and a concordance to the eight litigation categories of the court digests. Terminological Divergence classifies five recurring forms of cross-language mismatch in valuation vocabulary and records that no statute surveyed uses a general abstract noun as a term of art for the scheme's effect on value. The Basis of Value and the Approach separates the kind of value an assignment requires from the approach used to reach it, and shows one set of comparables supporting two figures under two bases. The overview also records a fact about the issue-1 dataset: once the remodel that Issue 1 disclosed is included, it is exactly noiseless, generated by a linear function whose $60,000 remodel term accounts for the Comp H residual. The dataset is extended here to sixteen properties with measured remodel indicators, leaving issue 1's published attributes and prices unchanged, so that every discrepancy in the first movement's worked examples is bias or a failure of identification, never noise.