Math is Hard: The Effect of Calculation Difficulty on Sales Tax Salience

Abstract

In this thesis I examine the effect that the difficulty of calculating sales tax has on how often people incorporate it into their purchasing decisions. This is known as tax salience. I ran a behavioral economics experiment in which I examined how often subjects incorporated sales tax into their purchasing decisions by measuring how often they purchased a good when the presented (tax-exclusive) price was below their stated willingness to pay but the actual (tax-inclusive) price was above it; By comparing the relative frequency of subjects calculating sales tax across different tax rates with varying calculation difficulty, I could test whether more difficult to calculate sales taxes were less salient. The data indicate that more-difficult-to-calculate sales taxes are less salient; subjects incorporated sales tax into their purchasing decisions significantly less frequently with a non-integer tax rate than with either a round integer rate or a non-round integer rate. The data also indicates that the magnitude of the tax and the difficulty of rounding the tax rate also affect salience. Economic and policy implications are explored. I discuss how low salience taxes carry a smaller deadweight loss than high salience taxes. Since more difficult to calculate taxes are less salient than more easily calculated ones, they should similarly carry a smaller deadweight loss. I also explore the implications of tax salience being a function of tax rate, demonstrating that it presents an opportunity to raise tax rates while reducing deadweight loss. I ultimately recommend that policy makers adjust sales tax rates from integers to nearby non-integers (within 1% of current rates) to increase economic efficiency.