مقاله انگلیسی رایگان در مورد سیاست های فرار مالیاتی و تقاضا برای وجه نقد – الزویر 2023

Abstract

This paper analyzes the relationship between tax evasion and the demand for cash by studying the effects of two measures to fight evasion: accessing taxpayers’ bank data and imposing thresholds for cash payments. We study the effects of these policies in Italy, where visibility of bank data and cash thresholds were recently increased. We show that both significantly affected cash holdings, which grew by about 1.5 percent of the GDP. Using unique high frequency data on cash operations and exploiting regional heterogeneity in tax evasion propensity, we find that accessing bank data pushes regions with higher propensity to evade taxes to convert more deposits into cash. On the contrary, higher cash thresholds do not increase cash holdings more in these regions. We rationalize the findings with a simple model of tax evasion and payment choices, where cash and deposits have different degrees of privacy.

Introduction

Despite the increasing use of electronic means of payment, cash is still widely used.1 If compared to other payment instruments, a unique feature of cash is its complete untraceability, which is highly valuable for hiding taxable transactions and wealth.2 While the role of tax evasion in the demand for cash has been widely explored, studies on the effects of measures to fight tax evasion on cash demand are scant.3

Anonymity in payments is a feature inherent to the use of cash: it provides a greater degree of privacy than other means of payment. On the contrary, the ability of different types of money to protect privacy varies, given that any payment system can shield the payee’s entire information set or a subset of that information. In fact, in account-based networks, the payer’s identity must be identified (Brunnermeier et al., 2021). If individuals are sensible to the number of observers, a demand for payment privacy and anonymity emerges (Borgonovo et al., 2021). In this context, the demand for privacy has two main sources: (i) a demand by individuals involved in illegal transactions, who try to reduce the probability of being incriminated (Masciandaro, 1999, Ardizzi et al., 2014b), and (ii) a demand by agents simply in search for protection from external scrutiny, not necessarily avoiding legal sanctions (Kahn, 2018). Indeed, it partially explains the emergence of new payment architectures and cryptographic procedures used to protect privacy, and the fact that they are viewed as close substitutes of cash for illegal transactions (Hendrickson and Luther, 2022). The literature on money and privacy has mainly focused on these aspects, that is, on improvement in privacy protection that leads to efficiency gains in the demand for money (Kahn et al., 2005), and recently on the introduction of central bank digital currency (Garratt and van Oordt, 2021), which can manage different degrees of financial privacy or different degrees of anonymity (Keister and Sanches, 2021). In this paper, we study tax evasion policies that interact with and alter the relative level of privacy between cash and deposits.

Effects on regions with heterogeneous tax evasion propensity

After having secured enough evidence on H1 and H2, we test H3 and H4. Given that we cannot observe cash demand and tax evasion at the individual economic agent level, we exploit variation across Italian regions in terms of tax evasion propensity to see whether changes in bank data visibility and cash thresholds had a higher impact on cash demand in regions with higher or lower propensity to evade taxes. Finding higher effects in regions where tax evasion is higher (lower) would support the idea that higher cash demand following policy implementation is driven by evaders’ (compliers’) behavior.

Measures of tax evasion are notoriously imprecise, because of its intrinsic opacity. There are several ways to approximate tax evasion propensity and a long debate has been opened about how to measure unreported, non-observed, underground, illegal, informal, shadow, and unrecorded economies.32 Our goal here is not to contribute to this debate, but to find a reliable approximation that captures the main differences between Italian regions. A way to approximate tax evasion propensity is through the non-observed economy, for which an estimate at the regional level for the period under analysis is produced and used by reliable institutions in Italy, oppositely to other measurements. In addition, considering the non-observed economy allows us to capture cash demand of agents operating in the illegal and informal economies among the others (see the examples in Section A of the online appendix).