A few days ago, I proposed an idea to replace means-tested programs with flexible benefits. Part of the idea was to get rid of the various income thresholds and instead start with a fixed sum of flexdollars that “fades out” at a rate of 20 percent for each dollar that a household earns.
It occurs to me that this can be thought of as two separate proposals.
1. Replace all earnings thresholds in means-tested programs with a fade-out.
2. Consolidate means-tested programs into flexdollars.
Since my main concern is to lower implicit marginal tax rates on low-income households, maybe the simplest thing to do is to focus on doing (1) first, and leave (2) for later. Thus, for each means-tested program, replace all earnings tests with a fade-out rate of 20 percent, meaning that you lose 20 cents in food stamps for every dollar of income. So if a family of four with zero income gets $8000 per year in food stamps, a family with $20,000 in income would get $4000, and a family with $40,000 in income would get no food stamps.
In fact, food stamps work sort of like this. See A Quick Guide to SNAP. But the rules are more complex. And then you have Medicaid, welfare, housing subsidies, and Obamacare subsidies, all with different approaches to means testing. I think it would be pretty straightforward to introduce a “flat tax” of 20 percent for all of these means-tested programs. With Medicaid, perhaps the fade-out could be applied to the Federal subsidy given to states, and then it would be up to states to pass this through to individuals.
Comments welcome.
The end goal should be to make people independent. Or put another way, make room for other people who need help.
By way of humor, if you are spending $666/month ($8k/yr) at the grocery store on poor choices, you will be spending a lot of Medicare money if you live long enough to qualify.
It’s a slippery slope, but I wonder how we can pay suppliers of food as well as healthcare for outcomes instead of volume.
A table or chart comparing before and after would be very helpful.
Pre-Tax & FlexDollars Transfer Income
Net Tax and Transfer (Now)
Net Tax and Transfer (Kling-SNEP)
One question is the estimate of how much the guaranteed income would be. At the very bottom, for a family of four with medicaid, food stamps, section-8 housing vouchers, child credits, child-care subsidies like head-start, etc. what is the amount? I think it could be pretty large.
That gets into the question of how much maintaining a 20% effective marginal income tax would cost. Perhaps not too much, because most income taxes are collected from the highest levels of income (the bottom 50% only pay 2% of income taxes as it is with top AGI in the low $30,000’s, and average fed income tax rate of about 1.5%). Though that’s not considering payroll taxes.
So reducing the MRT’s might mean not less taxation, but more expenditure in terms of delivering greater amounts of money to more people who currently face high MRT’s, and some money to people who currently pay income taxes. It’s very redistributionist, and it might mean that everyone in the the bottom 60% would still be a net recipient of federal money and have a very strong incentive to preserve that regime, but also agitate for larger payments.
There is also the tricky question of immigrant eligibility. Can this possibly work or be sustainable in an open borders world?
My quick take is that while I agree it is an improvement over the current system, this would be relatively complicated.
You’d have a parallel currency which would be good for many if not all things, and in any case could be traded out for dollars at a reasonable discount. As you stated, there would be the problem determining which items could be sold for the parallel currency. The reason for having a parallel currency for merit goods is that taxpayers don’t want to pay for booze or drugs, but the fact of the matter is almost anyone can get actual dollars to spend on these things if they want to do so, so there isn’t much point in limiting the grant to merit goods.
The bolt-on catastrophic insurance plan is one size fits all and might not be best for all situations. If the $7,500 flex dollars per person is all that there is, and the insurance only covers spending beyond $30,000, then a poor person with moderate to high health care expenses seems hosed.
My suggestions would be as follows:
1) Abandon the paternalism with the income grant – just give all citizens a pure cash benefit. Make the payments relatively frequent, say weekly.
2) Rather than a fixed sum per person, set the benefit so that it equals the poverty level for the household, splitting the amount evenly among all the adults (e.g. since the poverty level for a household of two is $15,730/yr, two roommates would split the weekly payment of $302.50). This effectively puts all citizens out of poverty (though I’d incorporate Charles Murray’s idea of allowing child support to be drawn from this grant). This grant could also effectively fund the support anyone who needs mental help or to make refuge in a homeless shelter. Allow people upon inception to choose whether to opt into the grant or Social Security. All other transfer programs are shut down, the federal government gets out of education entirely (but also forgives federal student loans as a sweetener). Social Security naturally fades out as younger individuals would likely prefer to receive many years of the income grant rather than wait for Social Security.
3) Pair the income grant with a flat tax, exempting as little as possible. A higher rate than 20% is probably feasible – most middle class workers are already paying 32.65% federal (7.65% FICA, 25% income) at the margin.
4) Instead of catastrophic insurance, figure out a long term budget for what the federal government is willing and able to provide in terms of health care and provide the per capita present value amount in the form of a massive health savings account. That’s likely a high amount, as we’re already paying for Medicare, Medicaid, ACA subsidies, VA Health, the health insurance tax subsidy, etc. Let’s just say for the sake of argument that it is $250,000 per citizen. This would give individuals a war chest with which they could finance much of their own health care, and family/friends could provide quasi insurance type capabilities if someone’s account ran out (and real insurance of course could be purchased, and we might see some experimentation with lump sum event insurance along side traditional PPO/HDHP plans). It would also overcome the liberal critique against HSA/HDHPs in that most health care spending is done by sick people who blow through their deductibles/HSAs and allow market forces to touch even very expensive health care services such as cancer treatments. The downside is that there needs to be a way to pick which goods are covered, but it seems that this will have to occur unless the government is kicked out of the health care business altogether. The account value could be scaled up by the growth in GDP per capita each year, and the unused balances in an account at death could be split evenly between the government and heirs.
Hopefully this makes things simpler. Everyone pays the same tax rate (and so taxes should mostly be automatic in the eyes of individuals as payroll taxes are), everyone gets a check putting them at the poverty level and everyone has a health care war chest.
Since I’m likely already well beyond gyob, I’ll end here.
Arnold,
Though the idea of “flexdollars” certainly assuages the concerns of those who would like to exert their own preferences over the choices of others, I don’t see any other reason to continue with the concept over simply handing people cash, which I outline here and here. Another knock against flexdollars is that at some margin, they are not going to be exchange media for something that people demand. That sets people up to make financial transactions at very high interest in order to trade their token for something with more liquidity, and these are the people you’re trying to help!
Are you wedded to the idea of using flexdollars, or are you introducing them for political salience?
After writing this, I noticed Justin has the same point above.
1. Not a good idea – getting rid of earnings thresholds would rule out papers that use these discontinuities for identification. More seriously, I have never understood why there are not more fade-outs. I wonder if it may be a remnant of the days when computers were not abundant.
2. As other commentaters have pointed out, this is a surprisingly paternalistic suggestion coming from a libertarian. I am curious what you envision the process for defining and approving “merit goods” would look like.