Optimal VAT Threshold(s).

Value Added Tax (VAT) has become a major source of revenue for governments around the world, with VAT taxation raising about a fifth of total tax revenues both worldwide and among the members of the Organisation for Economic Co-operation and Development (OECD). A popular element of a VAT tax system is the existence of threshold: the requirement, that is, of businesses (taxable persons) to register for VAT after their (gross) turnover (from supplied goods and services) exceeds a given level. But their level varies across countries. In the EU, for example, all countries (with the exception of The Netherlands and Spain) have VAT thresholds ranging from Euros 10,000 in Greece and Finland to around Euros 114,000 in the United Kingdom. There is also significant variation in VAT threshold across non-EU countries. Thresholds vary in terms of the sector business they belong, with some countries applying distinct thresholds to the provision of goods and services (such as, in the EU, France, Ireland, Malta, Portugal, and, for China and Japan for non-EU countries). The economic rationale behind the existence of VAT threshold relies on the observation that the (gross) turnover distribution of businesses is such that a relatively small proportion of them account for a significant proportion of VAT revenues. Given capacity of resources available to revenues administrations, and the significant burden (relative to sales) that the VAT system imposes on revenue administrations and small businesses, revenue maximisation might dictate that some small businesses should be excluded from the requirement to register for VAT. But the variation of the threshold across countries shows that it is a very contentious issue, reflecting, partly, that revenue administrations are somewhat averse in excluding taxable entities from the tax base and, partly, lack of a systematic approach in thinking about the optimality of the thresholds. Thresholds create a 'notch', and a VAT threshold is no exception. The reason for this is that for businesses which are close to the threshold, their tax liability (and profits) changes discontinuously. The implication of this is that for those businesses who are close to the threshold (from above) the incentive to move to the threshold and 'bunch' with others is quite strong--one the one hand they reduce sales but on the other they save one compliance costs. This, from an economic point of view, is inefficient. But there is another inefficiency in production decision along the production chain, as the VAT these firms have paid on inputs cascades through the input choices, if they are non-VAT registered (this is against the VAT principle). Despite the importance of the topic, the literature has not dealt with (with notable exceptions to which we turn to shortly below) the issue of optimal threshold and there is no, to the best of our knowledge, theoretical work investigating optimality of VAT threshold along the production chain justifying, or not, the existence, from an optimality point of view, of multiple thresholds. Despite the prevalence of VAT threshold, and its implication for tax revenues, it is therefore very surprising that the issue has not attracted the attention it deserves in the Public Finance literature. Perhaps this neglect arises because of the analytical complexities introduced when the VAT chain is modelled. This is the focus of this paper: To investigate whether the VAT threshold across the chain of production should be uniform or not and so different sectors should be treated differently along the production chain. It is shown that there is a case for differential VAT thresholds. The optimal threshold is also derived and calibrated to Bulgarian VAT data. [ABSTRACT FROM AUTHOR]