The progressive politics of pricing publicly-provided products

Whenever I make the points summarised in this post, someone invariably counters with the following median-voter argument:

• A policy in which only low-income households benefit will not command majority support.
• Making the program universal will ensure a more stable electoral base.

This argument is not without merit, and it's pretty convincing in certain contexts. But it doesn't work in all contexts: there are several important cases in which the traditional progressive preference for universal programs leads to regressive outcomes.

Let's start with two important cases in which universal programs generate progressive outcomes: health care and K-12 education.

A NBER paper by Sherry Glied makes the point that the Canadian health care system is 'modestly redistributive':

Putting $1 of tax funds into the public health insurance system effectively channels between $0.23 and $0.26 toward the lowest income people, and about $0.50 to the bottom two income quintiles.

Half of the public spending on health care goes to the lower 40% of the income distribution. It's not exactly Eat the Rich, but it's better than nothing. And if the choice is between that and nothing (glances nervously at our cousins to the south), then I'll take it.

Similarly, lower-income families benefit disproportionately more from the public funds spent on the K-12 system, if only because upper-income families have a tendency to opt out of the public system. And to the extent that education ministries concentrate resources on problem schools, then the system will be even more progressive.

These are examples in which universality provides outcomes that are progressive. Not as progressive as targeted transfers, but since the benefits are spread across a broader base, there is less of a chance that the median voter will defect from the coalition. But this does not mean that universal programs are always and everywhere progressive. If that point is not immediately obvious to you, ask yourself if a publicly-funded universal program to provide free yacht maintenance would be a considered to be a progressive policy.

This brings me to two examples in which this fundamental error in logic has led any number of self-described progressives to advocate policies that concentrate spending at the top end of the income distribution, for no better reason than that they can be represented as universal programs:

• Free tuition: [1], [2]
• Subsidised daycare: [1], [2]

The difference between these programs and free yacht maintenance is one of degree: all three concentrate public funds on the top end of the income distribution.

If you're a self-identified progressive, and if you find yourself looking for reasons to explain why it's okay for rich people to receive the lion's share of public money, you should ask yourself why you're even trying.

Comments

[Mandos]

God, you aren't beating this drum again... But you are.

[Mandos]

The point is that the median voter argument is not *the only* argument, just as modern progressive politics recognizes more as class oppression than *just* economic class oppression, important as it is. Subsidized daycare is part of a program of wealth transfer from a masculine (not always biologically male) class to a feminine (not always biologically female) class that is indirectly associated with matters of straight-up economic class.

[Patrick]

For daycare, it would be fairly easy to adjust the system, I think. You could adjust payroll income tax deductions up or down proportional to income and the number of children in daycare. The daycare provider could issue a statement at the end of the year saying how much daycare you actually consumed and your taxes would be adjusted accordingly. Wealthy people would get dinged more off the top, middle income less, and low income not at all.

[Stephen Gordon]

The point is that the median voter argument is not *the only* argument, just as modern progressive politics recognizes more as class oppression than *just* economic class oppression, important as it is. Subsidized daycare is part of a program of wealth transfer from a masculine (not always biologically male) class to a feminine (not always biologically female) class that is indirectly associated with matters of straight-up economic class.

Good lord. How in the world do you get "... so it's a good idea to give free money to rich people" out of that? And again, why are you even trying to justify a policy that uses most of its budget to give free money to the rich?

[calmo]

Dang if I can find the drum beat so offending to Mandos, Stephen...and ordinarily I prefer anything to the straight forward march.

someone [meritable debater worthy of squashing unconditionally] invariably counters with the following median-voter argument:

* A policy in which only low-income households [fuggedabout those misfortunates not reaching the household bar --this is not the Salvation Army ] benefit will not command majority support [with apologies to JS Mill...and those high-schoolers who digested 'On Liberty' ].
* Making the program universal [like GM's goal: a chevy in every garage ] will ensure a more stable [twas a horse tradin joke!...seriously, don't sink into Chomsky's paws like this. We want a responsible, alert, caring population, not a difficult to revive, low maintenance, trouble free...essentially dead ] electoral base.

This argument is not without merit, and it's pretty convincing in certain contexts. But it doesn't work in all contexts: [Right. Like the context of moaning about lousy public policy of distributing hard-working taxpayers money to freeloading losers...who should feel lucky to have the chance to scrub my yacht. ]there are several important cases in which the traditional progressive preference for universal programs leads to regressive outcomes.

My previous submission was reported to have "unacceptable data", which I scanned for viruses...and spellin mistakes, but I leave it here...not thinking that I am banned, but that forces beyond my control are interceding. I am only so funny, it appears.

[calmo]

Ok, so they are napping. Furthermore....
my response to Mankiw's complaint about the misreportage of the tax burden falling heavily on the wealthy...drawing heavily on the Laffer Curve's lesson (the heuristics, people: that was a good answer Johnny, showing the rest of the class, some of whom could be smarter than Johnny, that guess was not it.)
Ok, so the non-Mankiw crowd is up to speed on the merciful realization that the merciless disparity affords the wealthy the opportunity to bloviate about just how high their taxes are and just how unfair it all is that so few pay so much.
Nobody takes this as seriously as Mankiw...expecting the poor to start looting so they too, can contribute to a more respectable, possibly even distinctive, if not distinguished, portion of the national tax revenues.
Submission #2.

[westslope]

Hmmm, I'll have to look at daycare financing models again.

Haven't cranked the numbers, but have often wondered if Employment Insurance (sic) benefits for pregnant mothers and young fathers tend to benefit the working rich.

The rich also benefit from tax breaks for hybrid automobile technology. (That beyond the fact that these programs constitute a clear signal to the market that "more is always better".)

Inexpensively provided public fish and game also tend to benefit the rich, and within populations of anglers and hunters, suprisingly small numbers of individuals take the vast majority of the Queen's fish and animals. I would guess that the 'median voter' has little or no understanding of how somebody else is enjoying her lunch.

[calmo]

Really? 2 in a row...
Ok, Mandos (not Womandos...as we take the hint) lights upon the plight of women...setting off a contagion of other fires (like the plight of aboriginals esp in Canada, yes?...but we could mention others...let's: the ho mo no mo campaigners. The minute the churches get behind these guys, kiss your ass good bye hoping to raise a straight family. And straight means conventional and conventional means status quo and that means me and my dough against you bum-grabbing wastrels...as calmo seizes the moment to cast this asspect of Public Policy in terms we all understand: consenting adult terms.)
Last (see 3 strikes and I'm outa here) thing:

The difference between these programs and free yacht maintenance is one of degree: all three concentrate public funds on the top end of the income distribution.

The yachts until recently, got bigger and fewer...do we ("moderate", "median voter" candidates, tolerant of damn near every little stupid thing) really expect policy changes to redress mo than the redressing? Before you clout me, when it does happen, do we see gracious apologies and decorum or something slightly more...bloody?
Ok, go you little #3 devil...

[Neil]

The issue I have, as it applies to the tuition argument, is that the idea of spending money on lower tuitions is to even out the family incomes of those who benefit from them. If we lower tuition today, then yes, more high income people benefit than low income people. But when tuition is less of a barrier, more low income people can attend the best schools, so the move is less regressive.

[Mandos]

calmo, my moniker is a reference to my long-standing Tolkien-geekery. I've been known as Mandos for 15 years now, at least.

[Mandos]

Stephen: Considering that the argument has been presented to you probably before you've had this blog, and more than once, I'm not sure it's fruitful to pursue it as you're obviously not listening.

Well...let me say one thing. Your view attempts to critique universality as a form of wealth transfer, in that as implemented, it sometimes has the counterintuitive effects that seem to turn economists on. But it's never been clear to me that wealth transfer as such is the *end-goal* of the modern progressive/leftist/whatever perspective. (Clue: it's also a means to a political end.)

[Stephen Gordon]

Oh, I'm listening, but I'm not understanding how you can expect someone to believe that an agenda of reducing inequality can be advanced by giving free money to rich people. It's either a cynical game of bait-and-switch, or bad analysis. I'm hoping it's the latter, so the remedy is to make certain facts better known.

You keep insisting there there's some overriding principle that trumps giving money to low-income households, but you never say what it is, or why it can't be achieved by giving money to low-income households.

[Nick Rowe]

I'm surprised that Frances hasn't waded into this part of the debate, on Universality! A few years ago, Frances and I did a brilliant paper defending universality from both left- and right-wing critics. We showed that "universality" (as we defined it) is a feature of an optimal tax/transfer policy, using bog-standard optimal tax theory.

Here's an early ungated copy: www.ciln.mcmaster.ca/papers/cc98/univ.pdf

I don't think our theory would justify free tuition, or subsidised daycare (what do you think Frances?). But it would justify universal child benefits, drugs, glasses, etc.

[Patrick]

Mandos: A program can be universally available to all while the taxation policy that pays for it is progressive. I think this achieves your goal (as best I can decipher it) of improving the lot of women regardless of economic class while not unduly benefiting the wealthy.

[Frances Woolley]

Daycare is different from the examples such as free glasses that Nick and I used in our universality paper.

The distributional impact of daycare subsidies depends entirely on the design of the program. Something like the federal childcare expense deduction is regressive because, unless the family's lower earner is in a fairly high tax bracket, it's generally better to pay cash and evade taxes than get a receipt (since the child care provider will usually face a higher marginal tax rate than the consumer given CPP premiums, clawbacks, etc). So it's regressive but - due to widespread evasion - largely irrelevant.

Quebec's $7 per day childcare program is somewhat more progressive but, to the extent that it favours people who are clued in enough to get their names on year-long waiting lists, again favours the better off.

I think the issue with the Quebec-style program is, with $7 per day childcare, it's completely rational for someone to put two children into childcare, costing the government $2,000 per month in subsidies (say), in order to go to work and get a job paying $1000 per month (say). This is not obviously a great use of government resources - unless the quality of care provided by childcare is substantially better than home-provided care. Unfortunately this issue tends to be so emotionally charged and polarized - who wants to believe that the choices they made harmed their child? - that we haven't had a really great public policy debate on the subject.

I think the message of Nick and my paper in this case is - if you want to help families with children, give money to families with children - don't worry about targeting it, or worry about whether or not children are yachts etc etc.