The problem is high implicit marginal tax rates on many people who are eligible for benefits from means-tested government programs. I think that a generic solution might consist of flexible benefits.
One approach would be to replace all forms of means-tested assistance, including food stamps, housing subsidies, Medicaid, and the EITC, with a single cash benefit. For this purpose, we might also think of unemployment insurance as a means-tested benefit.
The classic approach is the negative income tax. What I would suggest is a modification of the negative income tax, in which recipients are instead given flexdollars. These would be like vouchers or food stamps, in that they can be used only for “merit goods:” food, health care/insurance, housing, and education/training. One way to think of this is that it takes the food stamp concept and broadens it to include the other merit goods.
Flexdollars would start at a high level for households with no income and then fade out at rate of 20 percent of the recipient’s adjusted gross income. This “fade-out” would act as a marginal tax rate on income, so we should be careful not to set the fade-out rate too high.
Suppose that a household receives $7500 in flexdollars per member. Thus, a family of four with zero income would receive $30,000 in flexdollars. A family of four with $20,000 in income would lose 20 percent of $20,000, or $4,000, to fade-out, and hence would receive only $26,000. A family of four with a $50,000 income would receive $20,000. A family of four with a $100,000 income would receive $10,000. A family of four with a $150,000 income would receive nothing.
At the end of the year, unused flexdollars could go into flexible savings accounts. Tghese could be used for medical emergencies, down payments when buying a home, or to save for retirement.
There are two ways in which this represents an improvement over the current approach. First, it ensures that implicit marginal tax rates are low for benefit recipients. As it is now, people with low incomes easily can find that if they work they lose more in benefits than they obtain in pay. I think that is very corrosive, and I would put a high priority on restoring the incentive for people to work, while still giving them the means to meet basic needs.
The second benefit is that it gives recipients more flexibility and choice. Just as food-stamp recipients can decide for themselves what groceries to buy, flexdollar users can decide for themselves how much to allocate to housing vs. food vs. training.
One problem with a negative income tax or with flexdollars is that some families are needier than others, particularly with respect to medical issues. Someone with a lot of ailments and little in the way of resources will not have enough flexdollars to pay medical bills (remember that there is no longer Medicaid in this approach).
The solution I would propose would be to have taxpayers provide extreme catastrophic health insurance that kicks in if a household’s medical expenses exceed $30,000 in a year. For every additional dollar of medical expenses over $20,000, the government would pay 90 percent. For example, a household requiring $100,000 would receive $72,000. Of course, households would be permitted to obtain private insurance to cover lower levels of spending and/or to cover the remaining 10 percent of higher levels of spending. Overall, this idea bears some resemblance to the idea of “catastrophic reinsurance” that was floated about ten years ago.
I am thinking that we would eliminate Federal support for unemployment compensation. Instead, perhaps a private-sector form of unemployment insurance might emerge, and households would be able to buy this using flexdollars. If it turns out that nobody wants to spend their flexdollars on unemployment insurance, then that might be a sign that unemployment insurance is not such a great thing.
It might be best to phase in implementation. The first phase might be to fold in the EITC, food stamps, housing vouchers, and health insurance subsidies. Those are all programs that already take the form of cash or vouchers given to households. A later phase would be to replace Medicaid and unemployment insurance with flexdollars given to households. (Of course, if states want to continue to continue Medicaid or to provide unemployment compensation, without any Federal dollars to support the, they are welcome to do so. I doubt that would happen.) Another phase would be to wind down all forms of housing assistance, mortgage subsidies, Federal aid to education, training programs, Pell grants, and student loan programs, and replace these with flexdollars.
One challenge with implementation is in deciding which goods and services are eligible for flexdollars. Just as the food stamp program has to decide which groceries are eligible, the flexdollar program has to decide what counts as eligible medical services, housing services, and education services. Yes, that opens up the floodgates for lots of rent-seeking. If that gets really out of control, then it would be better to give people a straight cash benefit.
This is just a concept I am toying with. Criticism welcome.
Note that I once wrote an essay that I called The FlexDollar Welfare State that was not about an idea of this character. Instead, the essay criticized the George W. Bush Administration’s domestic policy initiatives. Actually, the best thing about the essay is the discussion of the oxymoron of “company benefits.”
What is interesting is that workers are not naturally suspicious of companies that pay “good benefits.” Apparently, most people believe that “good benefits” reflect generosity and sharing by the company, rather than a shrewd, calculated effort to save on compensation costs. My guess is that the people who see through the scam of “good benefits” tend to gravitate toward self-employment, which allows them to take their payments in cash and buy benefits themselves.
This runs into what I see as one if the major flaws inherent in a negative income tax: You need either a high marginal tax rate or a high break-even point. In your example, for a family of four to pay any taxes at all, they would have to make at least $150,000 per year, meaning that the entire tax burden falls on maybe a tenth of income earners. What kind of marginal rate do you need in the actual taxpayers to cover that?
It seems to me in these discussions that an opportunity to change work habits at the lowest end of the scale is always ignored.
Why not have a base cash (or flexdollar) transfer to each person/household that is not really quite enough to get by on? Then, have a rapidly escalating subsidy for ANY work that any member of the family does?
IE: If the teenaged son of a very poor family wants a new pair of shoes, and works for $1 an hour for a few days in summer, the subsidy is huge, maybe another $9 an hour. For a meaningful amount of work that teenager can go get his shoes. Today that teenager’s best avenue for getting those shoes is to steal them. I think we should change that.
I believe that your proposal (if I am interpreting it right) provides a base amount that declines on the first dollar of work, and that this is wrong. Instead I think there should be incredible benefits from doing the first few dollars per hour of work. This initial foray into the workplace should receive a very large subsidy. This would decline once the total compensation reaches “enough to live on”. As with your proposal it declines gradually to zero at some higher wage. This eliminates the extreme marginal tax rates of declining benefits but also provides a very good incentive to any very poor person or family to find ANY work for a meaningful amount of extra cash. Keeping the base benefit just low enough increases that incentive.
A big argument against the minimum wage is that it prevents the most poor, least skilled people from obtaining their first taste of the working world and the chance to develop the most basic work skills. In my view we should pay top dollar to ensure they take that step, and make sure the basic benefit is just low enough to encourage everyone at that level to do so.
I would even go so far as to have the companies buying the day labor, or paying lowest wages, to pay the benefit as part of hourly compensation and then have them receive the subsidy directly. This makes it trivially easy for a poor person to just walk up to an employer and start working, and to get good wages for it. Those low-wage companies would compete to offer the highest wage, and that would put a floor on the real portion of the total wage they paid. They’d go into the poorest neighborhoods to recruit, they’d hand out leaflets at the end of the school year at the poorest high schools. I think this would be fantastic.
If we’re being brave in our proposals let’s be really brave.
This would never work, I would pay you to “work” for me, but also give you a loan, and charge high interest, thus collecting as much of the subsidy as I can. If the subsidy is higher than the value of the work, it will be scammed.
I deleted my first draft… to be “charitable”.
Other Federal programs are subject to fraud and any program would need enforcement.
However, I think if you ponder how much a small landscaping company already stands to benefit with a system like this you’ll realize the absurdity of these workers being hostage to some criminal enterprise that secretly claws back all of the benefit. Companies would compete for workers quite a bit harder when the wages they actually paid were quite a bit lower than they are now. Not to mention how many new enterprises would be started to try to take advantage of these wages.
And on the question of the taxes, the upper-middle and upper classes are already paying $10 an hour or more for basic labor. Plenty of low-skilled labor is not even for sale because they won’t pay $10… but they might pay $8. The actual taxpayers would purchase a lot more low-priced services under this scheme. That would count a lot towards the cost in taxes to pay for it.
Good point – but it could be solved by catching some of these scammers with severe punishments when convicted like 3 months jail. Make it not worth it for $7500.
“if they work they lose more in benefits than they obtain in pay. I think that is very corrosive” — Do you have empirical evidence that this is problematic? Also I am not sure that strong incentives to work are appropriate for single parents with young children.
@Brandon Berg. Arnold actually has an interesting twist that may avoid the problem. It’s the “merit goods” angle. There isn’t pure substitution between income dollars and flex dollars. At some point, the marginal utility of the merit goods has fallen to a point where the utility loss of a flex dollar is significantly less than that of an income dollar.
Sure, you’d have to model this all out and do some experiments. But I think it has a lot of promise.
Perhaps the government should even encourage a secondary market that converts flex dollars to true dollars? That way we could see what was actually happening at the aggregate margin.
Kevin:
This already happens with food stamps. I recall reading an article about some depressed part of Kentucky. It said that when food stamps become available, many food stamp recipients go buy cases of Coke from some place cheap and sell them to someone else (often at a discount to what food stamps paid for them), to get cash for whatever.
Max
Yes, I’m aware of it. But the government typically frowns on this practice. I’m saying it should encourage the practice so that the discount is transparent. That way, we would know that in aggregate, a flex dollar was worth 63 cents or whatever. Movements in this price could be indicators of changes that we might want to investigate.
One big question – Can Income taxes be payed with flexdollars?
Also, does the income tax structure stay the same as now?
If taxes are not eligible, then the 100,000 income family will end out with something like 10,000 flex dollars, and pay 20-25,000 in taxes, and they would be facing a marginal tax rate significantly higher than now, because the flexdollar phase out combined with the income tax bracket.
If tax is payable with flex dollars it makes a lot more sense, but a lot of complication could be removed by making the whole thing a refundable offset flat tax – the refund could be in flex dollars, but tax is payable with them.
This basically works out to giving everyone a lump of flex dollars – say 10k per adult 6k per child (age dependant), and then taxing all income at a flat rate – say about 40%. With these example numbers the break even point where the rebate cancels the tax would be about 25k for a single, 40k for a parent + child. These numbers work out fairly close to what the current system (Canadian rates – it’s what i’m more familiar with) has as average tax rates by income bracket, but eliminates the spikes in marginal tax rates that can act as a trap.
I think there would be enormous pressure from producers to have their products be labelled as “merit goods” and thus be eligible for FlexDollars.
And in a lot of cases, they might have good arguments. Take clothing, for example. It’s not on your list, but basic clothing is a merit good. But at a certain point, clothing becomes luxuries. Or perhaps phones. Given that there is already a government program to provide people with phones, surely phones are a necessary part of modern life. Or cars. In many places, cars are simply not optional, and they are a major expense. A poor person might be perfectly rational in choosing to allocate FlexDollars towards a car. Or fast food places. Surely a single mother working two low-paying jobs should be able to use FlexDollars at a McDonalds to quickly feed her children.
Then perhaps the true non-merit producers will start a campaign: “Even the poor deserve a little fun!” Surely even a poor person should be able to go to the movies or a ball game once in a way. I’d lay even odds that something like this could gain popular support. Or perhaps someone will suggest that poor people should be able to go to the theatre. After all, culture enriches the mind. I don’t think the true vices will be able to be ‘merit’ goods, but I think many of the more tame entertainments could be labelled so after a sustained public relations campaign.
(Think of a scene where rich plutocrats enjoy a movie, while a chain link fence keeps out the poor people in the rain, forever looking in at the warm scene where people are having fun.)
After a while, what wouldn’t FlexDollars be useful for? Maybe alcohol and gambling. But given that FlexDollars will be used for everything else, may as well skip the whole thing and give money directly.
One of the other major disincentives of the modern system, besides forcing work, is to massively diminish the personal cost of having children, while punishing marriage at lower incomes. I think a system design that doesn’t address this flaw while attempting to fix things is pretty weak.
How about fixed income for adults only (kids not counted), and a 20% differential for family units, as opposed to the unmarried. Same basic structure, but without those two existing problems.
You could have a system with an additional $2.5k as cash for discretionary stuff. Morgan Warstlers’ system had some good ideas for rolling in a work deal.
Professor:
Nice proposal. I had been thinking about a system much the same. Here are some differences, which one could think of as criticisms, I suppose.
First, I would include private disability insurance premiums and possibly private unemployment insurance premiums in your concept of merit goods, allowing us to dispense with the public disability program and maybe the public unemployment assistance programs.
Second, I would resist allowing easy substitution between food, housing, and (especially) clothing on one hand (call them necessities) from various kinds of insurance premiums on the other hand (call them insurances). Insurances only have a contingent and delayed value, so people with the highest time value of money would be likely to blow the insurances subsidy on more living space and junk food. That’s fine if it is their money, but does nothing to alleviate public concern about their poverty when they get injured, disabled, unemployed, or whatever. Note that there’s not the same problem running the other way: you should be allowed to blow all your necessities money on insurances. Presumably, the government would put minimum requirements into place around insurances you can buy with the part of your flexdollars that are limited to insurances. For example, you can’t use it to buy a health insurance policy with a lifetime cap on treatment. Or you can’t buy a disability policy unless it includes a provision that entitles you to buy insurance for children born disabled. But you could add features that don’t comply using your necessities flexdollars.
Third, while I like the notion of having a single dollar value for all people, I suspect that it might be better to vary it by age — especially for insurances — so that senior get more. At the very least, I’d make it vary to some degree with the higher health costs of being old, so that seniors can afford the same level of insurance as they age. I’m not sure whether disability premiums would rise or fall with age. (The older you are, the less time you have to be disabled. But the likelier you are to become disabled.)
Fourth, I think education and training money would need to be in a third, dedicated bucket. The dollars going into this bucket should also be tied to age. If we are thinking about replacing public schools, then the young should get a lot more, since it will mostly be spent as it comes in (for day care, then school). If we are thinking about just replacing assistance to people of college age, then it should be focused on those ages. I’d then have a steady trickle, so that people can save it up (assuming you have a savings account, discussed below) and use it for a refresher course or professional education from time to time.
Fifth, I’m in two minds about your account that extra dollars go into. If it was government-managed, it seems dangerous. My thought was to simply add the total subsidy to the household’s income (which means it gets taxed at the marginal rate AND allow people to use the unused portion to pay any income taxes that they owe. But if you do that, then someone will refuse to buy anything and then get a cash refund at the end of the year, which undermined the whole purpose of limiting them to merit goods. So I am coming around to your idea. Maybe the accounts can be privately administered, like an IRA. There would be issues around what happens when someone dies. By making the accounts individual, you avoid issues around divorce, marriage, etc. I’d make them immune from recovery for indebtedness and immune from distribution in bankruptcy. I suppose the accounts could have investment features, so long as current-period payments (e.g. for premiums) come out before going into the account and subjected to the ups and downs of the market. But that just means that cash received into each account would need to be initially invested into cash for a month or so.
Finally, we’d want to allow various governments to top up the federal flexdollars. For example, we could do away with much of the Veterans Affairs Department by topping up the flexdollars for veterans. Or, if a state wanted to be generous with its taxpayers’ money, it could top up the flexdollars under whatever conditions it wanted to. This might allow simple programs focused on behaviors that prevent poverty: Graduate from high school and get $500 flexdollars or whatever. Giving flexdollars alleviates the concern that cash benefits would be misused. Opening up the accounts to contributions from charities might be something to consider, too. Essentially, these top-ups would have very low marginal cost other than the actual contribution because the flexdollar system would eliminate a lot of administration and grant follow-up.
So, I think I end up with three separate accounts for every person (insurances, edutraining, and necessities). They would be at private institutions. The government would pay into them periodically (monthly or maybe weekly). Other governments and charities could pay into them as well. Unused balances simply stay in the accounts for future use.
Max
My vote is that flex dollars be scaled by the amount of work that someone puts in week to week. If someone works a full-time job (40 hours a week), they are eligible to get 100% of the flex dollar benefits for their income bracket. If someone only works half-time (20 hours per week), they are eligible to get 50% of the flex dollar benefits for their income bracket. If someone works 0 hours a week, they are eligible to get 0% of their flex dollar benefits.
…But, here’s the catch. Anyone who is employed less than full time has the option to work government-funded public works projects in order to top off their 40 hour work week, including those who are otherwise unemployed. This gives taxpayers some form of a return on their publicly funded hand-outs, reduce the incentive to remain unemployed, potentially build marketable skills for those that may not currently have them, and provides a market-based approach to raising the minimum wage (via commercial competition with the government for lower-wage workers)
I like your flexible benefit plan but a basic income guarantee is simpler and people can trade food soap etc. for drugs and alcohol with a little effort.
On healthcare I like your plan but I like
this better.
The state would provide insurance to all Americans but the annual deductible would be equal to the family’s trailing year adjusted income minus the poverty line income (say $25,000 for a family of 4) + $300. So a family of 4 with a trailing year adjusted income of $30,000 would have a deductible of $5,300. A family of 4 with a trailing year adjusted income of $80,000 would have a deductible of $55,300. Middle class and rich people could fill the gap with private supplemental insurance but this should be full taxed. This would encourage the middle class and rich, who are generally capable people, to demand prices from medical providers and might force down costs. They could opt to pay for most health-care out of pocket while the poor often less capable would be protected.