George Mason economist and blogger extraordinaire Tyler Cowen sends me this paper published in National Tax Journal by Michael Keen and Ben Lockwood. It looks into whether or not the VAT is in fact, as its advocates and detractors have claimed, a money machine.
The final column of Table 2 shows that governments in OECD countries have, indeed, tended to become larger after their adoption of the VAT, in the sense that the proportion of GDP taken in taxes and social security contributions was higher in 2003 than in the year prior to VAT introduction, by nearly six points.
So, yes, the VAT raises a lot of money. However:
The final column of the table shows that in most cases the increase in the overall tax ratio has been less than the revenue raised by the VAT itself. Thus, the revenue raised by the VAT has been to some degree offset by reduced revenue (at least relative to GDP) from other taxes. It will be seen in the next section that the nature and extent of such offsetting is of central importance in evaluating the money machine notion, and exploring this, controlling for other potential determinants of government size, will be a key part of the later empirical analysis.
The authors note that these offsets make the VAT a “weak” money machine. Their conclusion: The VAT in itself doesn’t appear to be, “at least in the purely statistical sense,” causing the growth of government. The whole paper is here.