I was surprised in the discussion of Anne Alstott’s No Exit to find so much enthusiasm for means-tested benefits which, I suppose, reveals more about me and the company I keep than about anything else. I am not completely opposed to means-testing: in some areas of policy, for example funding higher education, I think it can be an effective tool for benefiting the less advantaged. And sometimes it is, given the political constraints, the best that you can do in lousy circumstances. But as a general matter universal benefits are better, and more egalitarian, than means-tested benefits. I was going to write up a lecturely account of why, after that discussion, but fortunately got distracted by summery things like making Bakewell Tarts and hanging out with my kids. And a good thing too, because Shlomi Segall has subsequently published a nice brief account of the general reasons why people like me prefer universal benefits. I’ll add one thing that Segall does not emphasize: the perverse incentives of means-testing. So, for example, the UK government’s decision to rely on the means-tested Income Guarantee Support as a top-up for the state pension introduces a disincentive to save for those nearing retirement age who think they might need it; and the old AFDC in the US reduced dollar-for-dollar as recipients earned income; recipients faced an effective marginal tax rate of 100% which even lefties like me can see might be a disincentive to work. But Segall makes the rest of the case briefly, and has thereby saved me a lot of work (which I was evidently too lazy to do anyway).
{ 11 comments }
adam scales 08.22.04 at 7:03 pm
Well, perhaps you should go ahead and make that case, because I’d like to hear it. As a conservative, I readily admit that people like me are generally allergic to government benefits, and self-described “lefties” like you are more enthusastic. Let’s say, for one day, I became an enthusiast for redistribution. I cannot understand why my (new) allies on the Left are so opposed to means testing.
Does the Left really want to stand up and take a bullet for the mortgage interest deduction? There was an interesting debate last week over rents and habitability, and whether such rules raise prices or redistribute surplus. I believe that the M.I.D. does nothing but raise prices of homes, thereby pricing poorer people out of the mnarket and their “universal” entitlement to deduct mortgage interest. Wouldn’t it be better, if we have to subsidize people, to subsidize people who have difficulty buying any home – and NOT help out people who thereby get to buy a bigger home, or a second home? I can’t believe opposition to this is a LIBERAL position.
I understand many liberals favor universal services in order to maintain public support for them. I would always find this a morally questionable basis, but here I wonder about its empirical soundness. Do relatively wealthy people continue to favor universal services (say, social security) they will not need, because they are entitled to them? I am not expecting SS to be a significant part of my retirement portfolio, but I am willing to pay to help out retirees in need. Must this include ALL of them? Much of the retired population saw as their peak earning years the greatest sustained economic expansion in American history, and (at least a few years ago, and I doubt it’s changed) are the wealthiest segment of the population. Surely, there is some daylight between forcing Grandma to eat catfood and sending Ross Perot a check every month?
I don’t know whether poor people feel stigmatized by receiving “welfare” benefits, as the paper linked above suggests. I know I would, but I am not sure how bad that is. Certainly, I question whether it is worth spending untold billions more so that we can pretend that we are “all in this together.” I cannot believe that a wealthy SS recipient feels the same way as a poor one, and if that is true, then what, exactly, is the “dignitary value” of making the benefit universal?
One reason I am conservative is my skepticism that goverment benefits will generally be allocated effectively or fairly. Put me down with the “public choice” crowd. I don’t believe that public skepticism about more funding for education stems from the sense that “there is nothing in it for me,” bu t rather from cynicism about how effectively those public dollars are being spent. Don’t liberals want to concentrate government spending enough to fix the problems the spending is addressed to?
Assuming that there is no magic wand of “full funding” for whatever program of benefits is most important to you, I do not understand why it is preferable to “waste” money on people who don’t need it. Isn’t the price of universality mediocrity for, among others, those most in need?
John Quiggin 08.22.04 at 9:57 pm
Segall makes a pretty good case for universal access to public services, but I don’t find it convincing as regards monetary benefits.
Australia has always means-tested benefits and that doesn’t seem to affect popular support for them. The age pension, where means-testing is most relevant, is very popular, while supporting parents benefits, where the main issue is the absence of work-testing, are not.
As regards incentive effects, these are essentially the same for means tests and for income taxes. A correct analysis should look at the effective marginal rate of taxation arising from the interaction of the two. This analysis implies that clawbacks should be gradual (below 50 per cent marginal rate), but doesn’t generally rule out means testing.
Jack 08.22.04 at 10:03 pm
I’m certainly against means testing for most of the resons expressed here and also for universal healthcare and education but I do feel a little ucomfortable that this is inconsistent. If I send my child to a private school, why should I not get a contribution from the state equal to what the state would have spent had I not? That surely is a form of means testing. Why should the state not pay for private medical operations, at least at a rate it would pay in the state system? Means testing too.
Clearly i think the broader issues of enforcing a cooperative equilibrium come into play and ther e may be differences between teh case where only a minority opt out or where majority participation is required for effective functioning but the arguments about means testing on their own are incomplete.
On a different tack, I think means tests are often a form of fraud which allows governments to look generous and thrifty at the same time while actually screwing the people in the middle. There is a lot of this from new labour which has increased minimum incomes admirably and has not raised the top rate of income tax nd even cut capital gains tax. It has done this in part by raising the not particularly progressive national insurance, council tax and increased use of means testing. It seems that public debate does not have the bandwidth to deal with a third concept and can only trade off two things.
Matt Weiner 08.22.04 at 11:17 pm
the old AFDC in the US reduced dollar-for-dollar as recipients earned income; recipients faced an effective marginal tax rate of 100% which even lefties like me can see might be a disincentive to work.
This is not quite relevant to the main point, but this was not true during the entire existence of AFDC; at some point there was a “thirty-and-a-third” rule that allowed recipients to keep the first $30/mo. and a third of the rest of what they earned. According to this paper “thirty-and-a-third” was instituted in 1967.
Christopher Jencks, in Rethinking Social Policy (p. 89), says that thirty-and-a-third was eliminated in 1981 by the Reagan Administration (and the divided Congress), apparently on straight-up cost grounds.
Robin Green 08.22.04 at 11:54 pm
When I was on the dole here in the UK a few months ago you, the rule was the state claws back all your Jobseeker’s Allowance except some fixed amount if you declare that you are working part-time – I think it was five pounds.
Brett Bellmore 08.23.04 at 2:11 am
As questionable as “robbing Peter to pay Paul” is, it’s remarkable to find someone defending “robbing Peter to give Paul something he could afford to pay for himself”. It’s always a suprise to be confronted with just how contemptuous “liberals” really are of property rights.
Isaac 08.23.04 at 3:08 am
On instrumental grounds, means-testing is frequently better because a) it sharply reduces the bureaucracy necessary to carry out the program b) reduces the stigma attached to receiving such benefits and therefore allows those whom it is supposed to benefit to benefit c) makes it appear more fair (although, in effect, it is not: to give everyone a benefit requires funding and that funding comes from higher taxes, so you will pay but it will be done progressively on the other end of the income spectrum).
I’m surprised that people in thrall of the notion that government is incompetent can suddenly place such great faith in a means-testing bureaucracy…
abb1 08.23.04 at 8:33 am
On a case-by-case basis – considering all the arguments against MT vs. the total cost of entitlement in question. I am sure there are some vital programs that are simply unfeasible without MT.
Robin Green 08.23.04 at 1:08 pm
On instrumental grounds, means-testing is frequently better because…
I’m surprised that people in thrall of the notion that government is incompetent can suddenly place such great faith in a means-testing bureaucracy…
I think you meant to start your comment with “On instrumental grounds, universal provision is frequently better, because…” – no, Isaac?
Decnavda 08.25.04 at 12:29 am
Means testing is a bad idea for all benefits, even, hell espescially, cash, if you are a person who cares about either of two things: the progressivity of taxes or the negative incentive effects of taxes. If you care about the progressivity of taxes, what you want is for higher incomes to pay higher tax percentages. If you care about the negative incentives of taxes, what you want is the lowest possible highest rate.
Practically, and morally if you are any king of consequentialist, a means test IS a tax. Whereever on the income scale you place the means test, you raise the tax rate by the amount of the phase-out. Surpose we have government benefit X, and our two options are to A) have a one third phase out somewhere between Grandma and Ross Perot, or B) just send a check to everyone, including Perot. We have progressive income tax rates, and will have to raise them uniformly to pay for the benefit. The tax rate increase is higher for option B than for option A. But under option A, the tax rate for those in the phase-out range is increased by 33%. Result: Under option A, Perot doesn’t get a check, but he has a lower tax rate than those in the phase-out range, while under option B, Perot gets a check, he pays higher rates than everyone else, but his rate is less than the rate for those in the phase-out range in option A.
The more universially useful a benefit is, the greater the disincentive in the phase-out range because the wealthier are less tempted by market alternatives. If the wealthy are not allowed to use the park gym for free, they may not miss it due to their spending time on other amusements. But they have to eat, and could supplement their grocery shoping with food stamps just like everyone else. And everyone can use more cash.
(NOTE: This analysis does not apply to phasing out variable benefits that generally help upper incomes more. The best example is the American phase-out of itemized deductions, which includes the mortgage intest deduction discussed above.)
Jason McCullough 08.29.04 at 8:53 pm
I don’t know about the philosophical arguments, but here in the US opposition to means-testing is one of the big-dividing lines between the New left and Old left. The reason? The Old left has looked at the disposition of the FDR programs (AFDC, social security), and concluded that means-testing would turn social security into “welfare” and destroy its political coalition.
Kaus dealt with this line of reasoning rather handily in The End of Equality, I thought. Back before he went bonkers…..
Comments on this entry are closed.