Simple taxes are often spoiled by the complexities of tax incidence. We think we’re taxing consumers, and somehow businesses bear the cost. We think we’re taxing capital owners, and somehow workers bear the cost. It’s as dizzying as Three-card Monte.
Ian Roxan bravely enters the tax-incidence fray with his recent article on the value added tax (VAT). I must admit up front that the article delved into details of European law that are beyond my ken. I forged onward nonetheless, armed with moxie and the assumption that the general principles underlying American and European tax law are roughly the same. I maintain this assumption of shared legal principles despite the vast cultural gulf between Americans and Europeans. See, e.g., the metric system, mayonnaise on French fries, and castles.
The starting point for Roxan’s discussion is that the VAT is meant to burden consumers, not businesses. This is in contrast to the corporate income tax (CIT), which burdens corporations (more accurately, capital and labor) by taxing corporate profits. However, as Roxan explains, VAT-exempt products and services throw a monkey wrench into these VAT-incidence plans. Where an item is VAT exempt, the producing company cannot credit or deduct any VAT that was paid on inputs. The result is that the inputs’ VAT imposes a cost on the exempt business. Although this cost can sometimes be passed onto consumers through higher prices, in some cases it cannot.
Roxan explains how the exemption of holding companies’ financial outputs can mean that corporate groups bear the cost of the VAT, upending the tax’s assumed incidence on end consumers. Consider, for instance, a corporate restructuring, which involves various VAT-covered legal, accounting, and consulting fees. Where the ultimate end result is the issuance of VAT-exempt corporate shares, the corporation will not be allowed to deduct the service fees’ VAT costs. In this way, the VAT can burden certain inter-firm transactions within corporate groups.
To explore this issue in depth, Roxan does yeoman’s work of summarizing a line of European Court of Justice (ECJ) cases dealing with holding companies that pay dividends to their parent companies. He starts by describing the Polysar case, in which Polysar Investments, a holding company, paid various VAT-covered service fees, distributed dividends to its parent company, and then applied for a VAT refund. The ECJ held that Polysar was not entitled to the refund because it did not actively “exploit” intangible property, and thus was not a “taxable person” entitled to a refund. Polysar did nothing more than receive dividends, which is merely incidental to share ownership and therefore does not qualify as “exploitation” of its shares. Roxan notes that this holding is strange because it results in different tax treatment -- denial of VAT refund for business inputs -- merely because of the insertion of a holding company into the corporate structure.
Roxan then goes on to describe subsequent cases addressing the same question of VAT refunds for holding companies, all of which came out differently. Importantly, the holding companies in the subsequent decisions all provided various services to their subsidiaries in exchange for fees. From these cases, Roxan distills the following rule: Merely owning subsidiaries’ shares is not economic activity sufficient to allow for VAT refunds, but management of subsidiaries is considered economic activity as long as it involves fee-generating services. In other words, by adding fee-generating services, holding company expenses are converted into overhead management expenses of the whole business, enabling VAT deductibility.
Although perhaps easily gamed, Roxan notes the benefits of this quasi-bright-line test, deciding that it is better to be over-inclusive (that is, grant VAT deductibility more often) rather than under-inclusive. This position brings us back to one of Roxan’s main points: When holding companies cannot deduct the VAT paid on corporate inputs, the VAT burdens the firm, which it is not supposed to do. And, specifically, it burdens firms organized into corporate groups, inefficiently incentivizing certain business forms over others and undermining the goal of neutrality.
While this discussion may seem primarily of interest to non-US taxpayers, it is quite relevant to design of the U.S. foreign tax credit (FTC). The FTC rules allows a credit for foreign income taxes only. As a consumption tax, the VAT is noncreditable. But, if the VAT and the CIT both burden corporate profits, they are closer together than many assume. If VAT incidence overlaps with CIT incidence, perhaps the VAT should be partly creditable under the FTC? (Jordan Barry and I have explored this point elsewhere.) Roxan explains other ways that the VAT and CIT may overlap more than we typically think, including the fact that CITs often allow accelerated depreciation (alleviating tax costs on investment returns), and that a CIT combined with a payroll tax of the same rate is economically similar to a VAT (since both burden labor).
If the VAT and the CIT are closer together than we think, perhaps the same is true of Americans and Europeans -- castles and dipping sauces notwithstanding. After all, whether as consumers, workers, or capital owners, we all pay taxes.