Behavioural Marginal Propensity to Consume with Applications to Tax Cuts and Interest Rates

G. Charles-Cadogan

Working Paper

Abstract

We find that the Duesenberry-Friedman marginal propensity to consume (MPC) gap is driven by myopic  loss aversion (MLA) with Friedman’s MPC as reference consumption, and that that drives a consumption ratchet for intolerance for decline in consumption. We provide several applications of the theory. For  example, we prove that the limiting values for behavioural MPC (BMPC) and consumption-based MLA  (CMLA) index are 0.372 and 2.72, respectively, and show how those values help explain the difference  between expected and observed spending of payroll tax cuts in consumer demand surveys. Other applications include solving a Mirrlees type redistribution problem by proving that if tax cuts are used to  stimulate consumption, then a balanced budget supports transfers to the lower classes, reduction of tax burden on the lower middle class, and progressive taxes on the upper middle class and wealthy. Somewhat surprising, we find a consumption-based loss aversion channel for interest rates, and a government multiplier paradox in which the multiplier is only effective in Friedman-like negative tax regimes. Thus, policy measures aimed at low-income and financially fragile families might drive demand better than untargeted stimulus payments. We also find that the behavioural interest rate is weakly correlated with inflation, it responds faster than the short rate when inflation exceeds the 2% target, and DMs react to  shocks to inflation by becoming more loss averse and revert to normal after about 2 years. For every 1% increase in the CMLA (fear) index, the short rate decreases by 1-basis point to stimulate the economy.

Keywords: myopic loss aversion, marginal propensity to consume, negative tax, behavioural interest rate

JEL Classification Codes: C02; D03; D11; E21; G38, H22