I. The problem: the theoretical prediction of tax shifting and incidence versus its actual measurement, 429. — Practical importance of the latter in administrative law in tax refunds, 430. — II. Limitations of the field: definitions, 433. — III. The law on tax burden shifting, 436. — Precedent for recognition of tax shifting, 437. — Absence of precedent for its measurement, 438. — IV. The economics of the measurement of tax burden shifting, 440. — Refund provisions of the 1936 Revenue Act, 441. — Economic precedent for analysis of margins, 443. — Business precedent in use of spreads, 446. — Implications of use of margins in measurement of tax burden shifting, 449. — Asserted impossibility of proof, 451. — Conclusion, 454.

Introduction, 36. — I. Controversy concerning the nature of statistical demand curves, 37; Distinction between "time" and "dynamics, " 39.— II. Different types of "demand" relationships, 42.— III. Statistical determination of demand curves, static conditions, 43; Simple dynamic conditions, 43; First and second order dynamics, 44; Trend-ratios, 45; Link relatives, 46; Different phenomena measured, 46; Distortion by link relatives, 47.— IV. The scatter-diagram — choice of curves, 49; Demand curves for individual years, and shifts, 53.— V. Correlated changes of demand and supply, 54; Dealers' and speculators' demands, 55; Pigou's method, 56; Douglas' method, 56; Leontief's method, 57; Mrs. Gilboy's method, 58; Multiple and partial correlation methods, 59.— Conclusions, 61.