Abstract

Fisher’s Price Expectation Effect: an explanatory approach to Gibson’s Paradox?

Gibson’s paradox posits a positive correlation between long-term interest rate trends and the general price level. I. Fisher explained this fact with the price-expectation effect; according to this hypothesis, on entering into a credit relationship in an inflation, creditor and debtor agree on an inflation margin added to the interest rate, which is equivalent to the expected rate of price increases. This thesis has recently found ever more widespread approval and there are already numerous empirical studies to confirm it. These studies commonly assume that economic entities derive their price expectations by weighting price-change rates that have already occurred, thus permitting use of socalled distributed lags. However, in the light of various indications derived from polls, etc., it seems very improbable that the formation of expectations by economic entities is in line with the extremely simple, restrictive expectation structures implied by the use of the various lag models. Furthermore, over and above the “agreement” between the parties to the credit contract alleged under the price-expectation hypothesis, an attempt is made to specify modes behaviour which may lead to an interest rate increase during an inflation. For the FRG, however, a confrontation of such modes of behaviour with empirical data provides possible pointers in this direction only for the most recent period, in which connection, moreover, the effects on the interest rate are not always unequivocal. With regard to the empirical tests, it is shown that both partial estimates of an inflationary component in the interest rate and tests with comprehensively specified interest rate models are problematical for various reasons; moreover, the price-expectation effect cannot be clearly delimited statistically in relation to the rival hypotheses designed to explain the Gibson paradox, especially the income effect and the response effect. The author therefore comes to the conclusion that the price-expectation hypothesis cannot be counted as yet among the economic hypotheses of demonstrated tested significance.