In theory, the EITC is a simple program. But in practice — and in particular from the vantage point of recipients — it’s opaque and complex. It’s almost surprising how much recipients did know about how their behavior related to the refund.
Salam’s piece has many useful links for what I am calling the Setting National Economic Priorities project. One of the project’s ideas is to introduce a standard “fade-out” rate of 20 percent for all means-tested programs in the safety net. A next step might be to consolidate all such programs into a single “flexdollar” benefit program, which I have described in previous posts.
My priors, which I think are supported by the research cited by Salam, is that trying to use a program like the EITC for social engineering is a mug’s game. I think that the flexdollar idea is a reasonable compromise between offering a pure cash benefit and trying to do fine-grained social engineering.
Let’s assume the minimum acceptable lifestyle costs $30K a year for some particular family situation, even if one has zero income. That seems like a reasonable estimate to me. I’m assuming that flexdollars and 20% MRT until households before quite wealthy replace all payroll taxes.
Lifestyle must increase at a rate of 80% of any additional income, then everyone below a household income of $150K a year (above the 90th percentile) is a net recipient of subsidies, and only households earning over $200K a year are paying more than $10K in net taxes (which is already about the top few percent or so).
According to the Tax Foundation, in 2011, the top 5% paid $590 Billion out of total income tax revenues of $1,040 Billion, and they already pay average tax rates of 21%. Under the flexdollar approach, those average rates would need to nearly double. That’s pretty radical.
How are the numbers wrong, or what am I missing?
In addition to the implicit 20 percent MTR on flexdollars for each dollar of income, the income itself would be taxed at its current rate. The bad news is that this makes the total MTR rate higher, but the good news is that it is not the budget-buster you would get if you kept the overall MTR at 20 percent.
When I look at this chart (HT: John Cochrane), a single mother in Pennsylvania starts with about $45K in flexdollar equivalents.
Until about $30K, those flexdollar-equivalents are taxed at about 50% with increased income. Afterwards, they are taxed at about (actually, slightly above) 100%.
But if you taxed them at 20%, you wouldn’t stop subsidizing people until household income hit $190K which is, again, in the top 5% (though probably 1% for single mothers).
From the 50th to 75th percentile, the household incomes split from $35 to $70K and pay an average federal tax rate of 7%. But at $70K, $45K of flexdollars reduced by 20% of income is still about $30K, minus the usual $5K in taxes, is still a subsidy of $25K. We’re already up to 14.4% of all federal income taxes forgone at $70K.
The point at which 20% of flexdollars balance income taxes is about $120K, households below which are responsible for about 30% of federal income taxes.
So, you will still have to nearly double rates for everyone with household income above $200K to pay for the missing income taxes and extra subsidies.