Douglas Walker has posted a thoughtful comment on yesterday’s post, titled “wealth tests are taxes”. I think the comment is important, so reproduce the gist of it here, and attempt to give a response that it deserves.
Yesterday’s post centred on the plight of a Detroit woman on Supplemental Security Income (SSI) who works part time and lives frugally in a low rent trailer park, but chooses not to save any of her earnings because if her accumulated savings were to become larger than $2,000 she would become ineligible for SSI.
Douglas objects to the fact that the woman is receiving SSI even though she obviously has an income in excess of that necessary to provide for her basic needs:
SSI and other benefits are not intended to transfer income. Their purpose is to provide some minimal level of living that society deems adequate. In the example given below the woman had a part-time low wage job which, when SSI was added, provided her with current purchasing power above what she must have regarded as an adequate level of consumption. So, rationally, she could have saved the extra income or in this case seemingly irrationally spent it on marijuana. Her SSI should have been lowered to match her own evaluation of her consumption needs.
Moreover, Douglas thinks that the $2,000 allowable savings is too generous – potential beneficiaries should be forced to spend all their savings before receiving SSI:
If people have saving in the bank why should others be asked to reduce their consumption and potential wealth to preserve the existing wealth of someone else? It’s not fair to the average taxpayer. Let people draw down on their wealth until they are truly in need.
Douglas is correct in stating that the purpose of SSI is provision for basic needs of recipients. According to the Official Website of the US Social Security Administration:
Supplemental Security Income (SSI) is a Federal income supplement program funded by general tax revenues (not Social Security taxes):
It is designed to help aged, blind, and disabled people, who have little or no income; and
It provides cash to meet basic needs for food, clothing, and shelter.
But Douglas is wrong to assert that SSI is not designed to transfer income. It is impossible for government to help low-income aged, blind and disabled people without transfering income to them from general government revenue, i.e. from taxes that are paid predominantly by those who are more wealthy. What Douglas objects to, I suspect, is not that transfers exist, but that they are too large and are given to those who, after transfers from government, end up with more than a subsistence income.
I also think that Douglas is wrong to assert that the woman in Detroit was “seemingly irrational” by spending her excess income on marijuana rather than saving it. It is true that she purchased an illegal drug (marijuana) rather than a legal drug (alcohol or tobacco) that might have been the choice of others, but her failure to save was a rational decision. Even a few dollars of additional savings could cause her to lose her entire SSI benefit.
To discuss the plight of this SSI recipient further, it is helpful to look at some numbers, even if some of them are necessarily semi-hypothetical.
The 2013 Federal benefit rate for SSI is $710 a month. Some states supplement this (from ten to two hundred dollars), but Michigan does not. Some wealth is not counted in determining eligibility for SSI. Ownership of a primary residence, for example, does not count, nor does ownership of a car if one is needed because of lack of public transportation. Always included in the $2,000 wealth test are cash, bank deposits, stocks and bonds. I do not know how the SSI manages to monitor the amount of cash stuffed in mattresses, but apparently the agency is able to carry out this task!
Once an applicant passes the wealth test, an income test remains.
Earned Income is wages, and net earnings from self–employment. The SSI benefit is reduced by $1 for every $2 after the first $65 of earned income each month.
Unearned Income is all income that is not earned, such as income from pensions (including Social Security pensions), State disability payments, unemployment benefits, interest on savings, and cash from friends and relatives. SSI benefits are reduced by $1 for each $1 of unearned income, after a free allowance of $20 a month.
Deemed Income is the part of the income of a spouse with whom an applicant lives, which is used to reduce the amount of an SSI benefit if the spouse is not eligible for SSI. This does not apply to our Michigan example, as the lady lives alone.
Now for our semi-hypothetical example. Let us, for simplicity, disregard unearned income, and assume that our SSI beneficiary works half-time (20 hours a week) for the minimum wage, which is $7.40 an hour in Michigan. Her monthly income, then, would be as follows:
$641.33 (Gross wages)
-65.00 (Not counted in the income test)
$576.33 divided by 1/2
=288.17 (Countable income)
$710.00 (SSI Federal benefit rate)
-288.17 (Countable income)
=421.83 (SSI Federal benefit)
With these assumptions, the woman’s gross monthly income is $1063.16. Her net income would be less, as she would have to pay a 6.2% social security payroll tax on wages as well as a sales tax of 6% on most purchases of goods and services (presumably not marijuana, though!).
Let us suppose that the $710 SSI Federal benefit rate is a good measure of the cost of living at a basic level. Douglas proposes that we reduce the SSI benefit so that our beneficiary has only this amount of total monthly income. How can this be done? There are two possibilities, and neither is viable.
One option is to reduce the Federal benefit rate by roughly half, to $356.84, leaving the income test unchanged. This would yield the desired gross income of $710 a month for our lady, who has a part-time job, but other recipients of SSI – who are unable or unwilling to work – would be left with incomes far below that needed for survival in Michigan. This option is clearly unacceptable if our goal, in the words of Douglas, is to ensure that everyone enjoys “some minimal level of living that society deems adequate”.
A second option is to retain the SSI benefit rate of $710, but reduce it by one dollar for each dollar of income from any source, earned or unearned, with no exempt (‘not counted’) amount. Our semi-hypothetical Detroit beneficiary would find her SSI benefit falling from $421.83 to $68.67 and her gross income falling from $1063.16 to $710.
But would our lady continue to work 20 hours a week at a low-paid job? People – even those who are not wealthy – respond to incentives. Why would she continue to work if she receives no reward, no increased income from her sacrifice? This option, like the first, is clearly not viable. Its unintended consequence is the creation of poverty traps, from which the poor have no incentive to
exit. Wealthy taxpayers, though, could then point fingers at the ‘undeserving poor’ who prefer to remain on the dole rather than work for a living.
Is there a third option? I do not see one, but perhaps Douglas has one in mind.
Tags: targeting
Many thanks for the comment on my comment.
Let me mention I should have put forward ‘a limine’ my concern is not so much with the technical questions surrounding means tests as the implicit assumptions underlying discussions of this subject. I accept that means tests have all sorts of unhelpful disincentives to work and save and alluded to this in my comment. Let me also say my comments on income transfers were directed at able-bodied beneficiaries and not the elderly poor and disabled who I believe are ill served by the present system. Perhaps I should have been more explicit about the context of my comment.
I also mention I do not have any problem with a welfare beneficiary having a “rainy day fund” of $2000. Depending on individual circumstances this might be prudent. My comment in this regard should been seen as directed at those beneficiaries with a large capital account well beyond their current needs and/or those who use income transfers to further accumulate wealth.
Of course it is impossible to help poor people without transferring income to them. My comment here is not about the process but the purpose. My concern is the unstated assumption that a transfer to the poor is treated as manna from heaven with no effect on the average taxpayer providing the transfer. Of course it has. Money transferred to welfare recipients and used for consumption beyond their basic needs (trips to Disneyland and the like are not unknown) or to maintain and augment wealth positions will have been taken from people who could have better purposes for the money they paid in taxes than the able-bodied poor.
While it may be true taxes are paid predominantly by those who are more wealthy, it is also a meaningless statement as a matter of public policy or even basic morality. Average and even below average income households also pay taxes and are in effect transferring part of their income to the poor and to them the transfer is not insignificant. It may well be that a taxpayer is in the top 5 to 10 per cent of income earners. At the same time even the rich may well have heavy expenses important to their families such as tuition for their children, a need to accumulate saving to buy a house or a million other good reasons the tax they pay negatively affects their welfare. The blanket statement that “the rich can afford it” is nonsense both because not only the rich pay the tax and because taxes on the rich may have high opportunity costs to their households.
To clarify the point about transferring income: On the policy level there is a difference between a transfer intended to provide generalized purchasing power to the poor for any purpose and a transfer intended to provide a poor recipient with some minimal but socially acceptable level of living. In the first case, which we might refer to as “an unrestrained transfer of income”, saving — even using the grant for a trip to Disneyland — is allowed as a matter of policy because what is being transferred in generalized purchasing power for recipients to use as they wish. The second case, which we might refer to as “a transfer of resources in support of an adequate level of consumption”, would restrict how the transferred income is utilized because the purpose of the transfer is targeted and paying taxes has real opportunity costs to all taxpayers, and the tax burden should therefore be minimized.
Discussions of means tests in the academic literature presume an unrestrained transfer of money income. I suggest policy regarding support for the poor was originally intended to provide an adequate level of consumption and should be analyzed as a restrained claim on the taxpayer. In this case the question of the effects of means testing is secondary to what the level of the transfer should be and minimizing its cost to the taxpayer. Moreover, since income transfers are not financed by manna from heaven any discussion of means testing should address possible effects on the average taxpayer. As I mentioned in closing my comment, economics is about tradeoffs and scarce resources. There was no discussion of tradeoffs between the recipient and the taxpayer in the Tdj. On wealth means tests, one might well ask: Why should the average taxpayer reduce their consumption and saving so that a welfare recipient could maintain or augment their capital account? As a policy matter this is the question that should be answered.
Finally, I mention I have no comment on the example. How SSI or any other program is implemented and managed is beyond the scope of my comment. How much the poor should receive and how to deal with the complications of a terrible welfare system are beyond my comment and indeed the Tdj itself. I readily acknowledge the present system leaves many poor people in a condition of genuine want, the many programs now in place have perverse incentive effects, and efforts should be made to reform it. But these are questions of personal opinion and issues of politics to be weighed by legislative bodies. I make no comment on them here.
Thanks for the clarification, Douglas. I was pleased to see you agree that means tests have adverse effects.
But I have two questions.
First, how do we distinguish the able-bodied poor from those who are not able-bodied? This would require another test in addition to (instead of?) income and wealth. I suggest “instead of” since we agree that income and wealth tests have very bad effects on incentives for work and saving.
Would you restrict SSI benefits to those who are totally blind or paralyzed? If not, what degree of physical incapacity should qualify an applicant, and how can we measure it?
Second, how can we determine on what goods and services taxpayer money is spent? The Detroit woman in our example could argue, with conviction, that she was purchasing marijuana with her own earnings, and using her tiny SSI pension only to purchase basic necessities.
Mexico City provides all residents from age 68 with a universal pension equal to 15 days of minimum wages each month. The pension is credited to a debit card that is in the possession of each pensioner. The card can be used at any of several supermarket chains, but only to purchase food and health care products. Purchase of tobacco products and alcoholic beverages is blocked. This makes taxpayers feel good, because government money is not used for purchase of unnecessary or harmful products. Since money is fungible, taxpayers of Mexico City are undoubtedly supporting vices, despite the attempt to prevent this. Using a debit card to purchase food allows a pensioner to purchase cigaretes and rum with cash that would otherwise have been needed for food and toiletries.