One could see the Quill/Wayfair issue as helping to illustrate my old point that tax cuts or tax preferences can make government effectively “bigger” even if they reduce tax revenues, and thus that their elimination may effectively make the government “smaller.” At least when we are talking about fiscal matters – let’s leave aside for now, say, the issues around government agents who put children in cages – a meaningful, rather than formalistic, view of the “size of government” should be based on its distributional and allocative effects, relative to some baseline. (Although the choice of baseline is admittedly a vexed issue.) Thus, suppose that in Case 1 the government “taxed” $X from you on Day 1 and gave the same amount back to you (as “spending) on Day 2. Versus, in Case 2, it took half as much from you on Day 1 but either gave it to someone else or spent it on subsidies for the coal industry. I’d say the government is “bigger” in Case 2 than Case 1, even though the formal measures of “taxes” and “spending” are lower.
Giving Internet sales an effective exemption from state sales taxes, against the background of general under-collection of use taxes, could be viewed in tax expenditure terms as analogous to taxing all sales and then giving the money back to Internet sellers as a special outlay on their behalf. The fact that the effective exemption arguably wasn’t intended as a subsidy is immaterial if the question we are asking is simply what level of distortionary economic effects result from state sales taxes. These effects may now be lower, and if the states want to have the same net revenue as before they can do so by lowering their rates. If they choose increased revenues, this might conceivably lead to “larger government” in some dimensions, but there would still be an offset by reason of the greater neutrality as between retailers.