Although locating a company in a tax haven is not illegal per se, it is likely to be part of a scheme purported to erode the tax base or to shift profits to less-taxed jurisdictions. For this reason, this type of location decision is usually targeted by anti-avoidance laws, that can take the form either of specific rules or general standards that, ex-post, sanction or limit the location decision. However, rules entail higher drafting costs and are easy to circumvent whereas standards entail more uncertainty costs. The goal of this paper is to illustrate that the risk of aggressive location decisions can be predicted ex-ante using publicly available data and that this prediction can be used by tax authorities. In the paper, we do two things. First, we use publicly available accounting data for the period 2015–2019 on 4031 group ultimate owners (GUO) of active listed companies resident in one of the 27 European Union countries to predict the probability that these companies would have at least a subsidiary in a tax haven, by spring 2021, as well as the intensity in the use of tax havens. Second, we discuss how this prediction can be used by tax authorities in the context of a new administrative preventive approach that complements the traditional legal approach. This approach can increase welfare by reducing uncertainty, thus increasing investments and economic growth.

尽管将公司设在避税天堂本身并不违法，但这很可能是旨在侵蚀税基的计划的一部分或将利润转移到税收较低的司法管辖区。因此，此类选址决定通常是反避税法的目标，反避税法可以采取具体规则或一般标准的形式，对选址决定进行事后制裁或限制。然而，规则的起草成本较高且容易规避，而标准的不确定性成本较高。本文的目的是说明可以使用公开数据事前预测激进选址决策的风险，并且税务机关可以使用该预测。在本文中，我们做了两件事。第一的，我们使用 2015 年至 2019 年期间关于居住在 27 个欧盟国家之一的活跃上市公司的 4031 家集团最终所有者 (GUO) 的公开会计数据来预测这些公司在避税天堂至少拥有一家子公司的概率到 2021 年春季，以及避税天堂的使用强度。其次，我们讨论税务机关如何在补充传统法律方法的新行政预防方法的背景下使用这一预测。这种方法可以通过减少不确定性来增加福利，从而增加投资和经济增长。我们讨论了税务机关如何在补充传统法律方法的新行政预防方法的背景下使用这一预测。这种方法可以通过减少不确定性来增加福利，从而增加投资和经济增长。我们讨论了税务机关如何在补充传统法律方法的新行政预防方法的背景下使用这一预测。这种方法可以通过减少不确定性来增加福利，从而增加投资和经济增长。