Getting the Numbers Straight on VAT Revenue
June 19, 2015
Yesterday, a CTJ blog post estimated that Paul’s VAT would likely raise about $1.1 trillion a year if fully implemented in 2016—far below the $2.3 trillion in annual tax cuts that would result from other components of Paul’s plan. We estimated that this would translate into a $1.2 trillion revenue loss in the first year—and a $15 trillion budgetary hole over the next decade.
The core claim to this blog post that we disagreed with was the $1.1 trillion dollar a year estimate, which is substantially different from ours (close to $2 trillion.)
The main point they make is worth talking about and addressing, though it’s ultimately an erroneous objection. They rightly point out that including government workers in the VAT base doesn’t have any net effect on revenues. The government, as the Tax Policy Center puts it, is simply paying tax to itself.
This is how the current system works, and the Paul plan would not change it substantially. The current income tax, as TPC notes, already includes most of the value-added in the public sector. If you think about a federal employee getting money withheld from his paycheck for federal taxes, we already see in practice, under the current system, the practice of the federal government taxing itself. (Alternatively, you can think about the federal government simply buying a good or service from a private contractor. They all pay taxes, too.)
The government taxing itself, in other words, is not a new “shell game” introduced under the Rand Paul tax plan, but just a fundamental aspect of how our tax system has always worked.
The reason VAT plans often include government is mostly for accounting purposes, such that the government bases its decisions on apples-to-apples comparisons where everyone pays the same taxes. Think about the tradeoff between a contractor and an in-house employee here. The true budget cost for a VAT-exempt employee is actually higher than the cost for a VAT-paying contractor, which would make things difficult to account for.
But I don’t want to dwell on this too long, because it’s not actually the core of our disagreement with CTJ. Government consumption is not the major source of our difference with CTJ.
CBO’s analysis showed what a VAT might raise if private spending on health care, education, and religion were excluded from the VAT base, as is commonly the case with existing European VATs and sales taxes in the United States. That, plus some other differences between CBO's analysis and TPC's, meant that the CBO's plausible VAT base clocked in at just 61 percent of total NIPA personal consumption.
We used CBO's 61 percent figure in calculating the tax base and potential revenue from a plausible VAT.
The problem with this is that in using that base, one would add in all sorts of exemptions and exclusions not specified in the Paul plan. It does the public no good to score a plan that no one proposed. The 61 percent figure comes from a CBO report from 1992(!) analyzing a potential credit-invoice VAT (the kind used in Europe.) The CBO preferred the credit-invoice method precisely because it allows for the sort of exemptions described above. Senator Paul's business tax is equivalent to the subtraction method, which makes those exemptions essentially impossible to create in the first place. It's much closer to being a combination of the payroll and corporate income taxes it replaces.
CTJ did score a plan with a 14.5% VAT, but one with little resemblance to what Senator Paul proposed.