payee of support or other payments, since most postwar experience has been with inflation, not deflation.) The greater the protection against unforeseen price changes afforded to the seller or payee by the protective clause, the greater is its cost as measured by what the seller or payee gives up somewhere else in the contract (although such tradeoffs may be less significant in the case of court-ordered agreements, such as child support). In the case of commercial contracts, the highest form of protection might be based on some index of the market prices of raw materials, and possibly the labor services, that the seller purchases. An alternative slightly less costly form of protection might be a price index of materials purchased or goods sold by the relevant industry. The most general, and least risk-free, form of protection is an index of some general measure of inflation. In that case the seller bears the “basis” risk of unforeseen relative price changes in the particular goods that are relevant to his costs but is protected against the common component of price changes manifest in inflation.
For purposes of inflation protection, the usefulness of the CPI, which covers only final consumption goods and services, has to be evaluated against that of available alternatives: the GDP price index, which covers all final goods and services, and the finished goods producer price index (PPI), or one of the PPI stage-of-processing components. Individual and family-related payments, such as child support, have some of the character of compensation, like the public benefit payments discussed above. In this case, the CPI may be the index of choice in protecting household consumption expenditures. However, for commercial contracts, the GDP price index may be superior if the broadest inflation protection is desired, while the narrower coverage of the PPI or one of its components may be preferable in other circumstances.
Most of the design issues that we have been considering in this report apply to the use of the CPI as an inflation index in individual and family compensation-type payments, and we do not discuss them further here. These design issues—for either the CPI or alternative indexes—are not critical for its use as an inflation index in business contracts, except perhaps during a transition period following the date when technical changes are incorporated in the index. Index design or measurement techniques may be periodically changed, with the result that the new index tends systematically to grow at a different rate relative to the prices of interest to a seller than it did in the past. An increased application of quality adjustment techniques in the CPI, for example, might lower its future rate of increase relative to that of the set of prices relevant to a seller. Such differences will eventually be taken into account by the parties to new contracts and reflected in the way contracts are written or in the cost of the insurance.
From the standpoint of sellers’ calculations, the relevant prices of purchased inputs (included in some PPI components but not in the CPI) are presumably themselves quality adjusted. Conceivably, the increased use of quality adjustments in the indexes might make the trend growth of the indexes track quality-adjusted input prices more closely than in the past. Other changes in the design of