The Government Does Not Own 100% of Everything

If I am reading this right, then Suzanne Mettler believes that any tax rate lower than 100% is "indirect social policy":

Conversely, many of our mostly costly forms of social provision today camouflage government’s role as a provider of social benefits. They do this by channeling benefits through the tax code, as does the Home Mortgage Interest Deduction, for example, and/or through subsidies to private organizations, such as employer-provided health insurance benefits, for which recipients are not required to pay taxes. These latter, indirect policies are the ones I call the “submerged state.” It is easy for citizens to miss government’s role in these policies, and to assume that only the market at is at work.

Now as it happens I am against deductions and all other loopholes as a matter of course. But I do not define "things not taxed" as the government acting as "provider of social benefits". Mettler, it seems, does:

The difference between the direct and indirect social welfare policies, however, is illusory. From an accounting perspective, they are the same thing: both impose costs on federal spending and add to federal deficits.

This is only true if you start from the assumption that 100% of national income belongs to the government, which then distributes according to a mix of "indirect" tax breaks and "direct" spending programs. I.e., it's only true if you believe that there is an implicit 100% tax rate, from which the government deducts differing percentages based on a number of criteria. In that case a tax deduction and a spending program are the same thing.

But if you start from the assumption that 100% of national income does not belong to the government then this makes no sense at all. If I buy a house and the government does not tax a small portion of the amount I pay for it -- I still pay taxes on the purchase of the home... I just get to deduct the interest -- I am hardly "getting" anything from the government. I'm just not having something taken away.

And in that case the only thing that adds to federal deficits are spending increases. Think of a scenario: there is a tax rate of 0%, and no federal spending. The government decides to pass a child tax credit equal to 5% per child. Does this add to the federal deficit? No, because 5% of 0% is 0%. Under this scenario the child tax credit is meaningless. Now suppose the government was to keep taxes the same but spend 5% on the child. The deficit goes up.

More:

In most cases, tax breaks distribute resources by permitting people to pay less in taxes, rather than by paying out dollars or providing services. That aspect of their design makes it easy to construe them as tax cuts rather than as social provision. But this, too, is a false distinction. “Social tax expenditures” assist people with particular circumstances, granting them resources to which others are not entitled. This is in stark contrast to across-the-board tax cuts to all Americans.

That's true. People without children cannot claim child tax credits. And the child tax credit is a politically-motivated tax exclusion designed to encourage better care for children (or, perhaps, having more of them). But it is qualitatively not the same as a spending program that is also intended to benefit children. The latter redistributes income from some people to other people. The former does not. The latter increases the budget deficit. The latter former does not. See? It's a very important distinction.

Note also that according to Mettler's definition a progressive tax code is a government "social provision" because the tax does not apply equally to everyone.

You could make the same argument about some of the other programs Mettler discusses. The GI Bill could be considered part of the compensation contract that the military makes with service members. I know my siblings in the military (there are currently five of them) think of it that way. Many wouldn't join the military just for the salary; the fact that part or all of their college education is included in the package is what tips the scales.

If you look at the famous table that Mettler produced in her paper, the programs that people tend not think of as government programs are things in which there is no redistribution happening. All of them are tax credits (except for student loans, which are paid back usually with interest). The things that people do think are government programs are the things in which there is redistribution happening. Mettler recognizes this:

In short, the fact that citizens often fail to recognize these policies as government social provision is attributable not to some fault of citizens, but rather to the characteristics of the policies themselves.

Precisely. And the characteristic of the policy is what is important. A policy of not taxing 100% of income is not a government intervention. It is the absence of a government intervention. A policy of not taxing mortgage interest is similarly a lack of taxing mortgage interest. People have different attitudes about whether these are "social provisions" from the government because they have different definitions of what constitutes a social provision from the government. Remember the outcry when Obama started talking about "tax expenditures" a few months back? That's because people immediately recognized what he was talking about: he wanted to raise taxes and redistribute the proceeds to others. So the phenomenon Mettler is describing is at least partially about semantics, not just cognitive dissonance.

I do not write this to disparage Mettler's research program, which I find very interesting and valuable. And the phenomenon she describes certainly happens sometimes. But I see the tendency to use the accounting that Mettler uses, which presupposes that 100% of national income belongs to the government, from a lot of political scientists. I've never understood it. It goes against all of the common principles pertaining to the nature and role of liberal government in a democratic society.