Scott Sumner  

"Public opinion" is not what you think

The Behavioral Econ of Paperwo... Trust Assimilation in the Unit...

Many people anthropomorphize the public, and talk about "the public" having an opinion, just as you or I might have an opinion. I view this as a big mistake.

Elsewhere I have argued that polls on policy questions are very misleading, and the results heavily depend on framing effects. Here I'd like to discuss another problem with so-called "public opinion".

Suppose you agree with me that the tax deductibility of health insurance is a huge distortion, which causes excess spending on health care. But polling experts tell you that the public likes the deduction, and hence nothing can be done about it. Should we believe them?

It would certainly be politically difficult, if not impossible, to immediately repeal the tax deduction for health insurance. But suppose Congress passed a law capping the amount that could be deducted. They did not do that, but they did pass the so-called "Cadillac tax", which has much the same effect. They also indexed the threshold to inflation, which is generally less than the rate of growth in health expenses. So the cap will gradually become more effective over time. (Unless it is repealed, which is possible.)

Now suppose that several decades from now a majority of insurance plans are up against the Cadillac tax threshold. At that point Congress replaces the tax deductibility of health insurance with a fixed tax credit for anyone with health insurance, equal to the average tax savings of the current system. In that case, the marginal cost of health insurance would rise sharply, considerably reducing the distortions in the health care system. But the public would still have their tax benefit.

I'm not saying this is going to happen, but it would certainly be far more politically feasible than simply repealing the tax deduction for health insurance.

News out of the UK provides an almost perfect example of the problem with public opinion. The Conservatives proposed eliminating a cap on how much any family was required to pay for nursing home care.

Launching her manifesto last week, Ms May said the Conservatives would ditch a planned overhaul of the social care system adopted by her predecessor David Cameron, which included putting a £72,000 cap on social care costs for any individual.

Under current rules, only the poorest people get state help towards their costs. Those with assets over £23,250 have to pay the full costs of care although for people being cared for at home this figure did not include the value of their house.

Mrs May's original plan sought to dispense with the cap and instead protect up to £100,000 in assets for those having to pay their full care costs.

Andrew Dilnot, author of the Cameron government's social care reforms, said last week he was hugely disappointed by Mrs May's proposals. He said it would leave people "helpless" until they were down to their last £100,000.

The populist press immediately began complaining about a "dementia tax" and indeed in a sense it would be a 100% tax on savings above that cap, for families with a loved one who needed social care. The firestorm of criticism forced the Conservatives to do a U-turn on this proposal, abandoning it despite a big lead in the polls (the election is just weeks away.)

Here's what I found interesting about this event. It seems to me that the Conservative Party's proposal would have also faced a firestorm of criticism in the US, if it were not already the law of the land. But unless I'm mistaken, the "dementia tax" proposal that was rejected by the British public is already the way nursing home care is handled in the US.

Now you might argue that the UK is different; they have the NHS and we don't. But their single payer system does not cover all nursing home care. I see this as a case of status quo bias. There are many aspects of our economic system that would never even be enacted here, if they were not already in place.

The art of politics is sort of like the art of boiling a frog. The key is to introduce sensible reforms so gradually that the public doesn't even know what hit them. Then a few decades later the public wakes up in an entirely different world.

This might sound undemocratic, and it would be if there were such a thing as "public opinion." But there isn't. If the reforms are enacted by democratically elected politicians, then they are democratic reforms.

The Cadillac tax is a dagger aimed at the heart of the medical industrial complex. The proposed elimination of the tax deductibility of state and local taxes is a dagger aimed at the heart of high tax (blue) states. These special interest groups are far too powerful to attack directly, and perhaps even these indirect attacks will fail. But they are probably our only hope for success.

Screen Shot 2017-05-22 at 10.25.45 PM.png

COMMENTS (14 to date)
John Hall writes:

Good post. I feel like there's a book to be written on small (marginal?) changes to policy that could have big long-run effects.

I would probably also note that U.S. immigration policy has followed a similar pattern. A small change to the rules in 1965 resulted in big changes 30-40 years later. God, I've been reading too much Steve Sailer!

Andrew_FL writes:

Not sure what point you think you're making with that chart appended at the end but health care expenditures tend to increase super-linearly with consumption levels and the US has basically the most developed consumption level in the world. So we'd expect to see a chart very much like it even if policy weren't in the thrall of the "medical industrial complex."

Thaomas writes:

What about gradually raising the amounts that are deductible for "retirement" savings to convert the income tax to a consumption tax?

Thaomas writes:

A semi-snarky, semi serious question. Unless state taxation is just a "bad thing" that should be minimized, why is it good not to allow a deduction (or partial tax credit) for state and local taxes? I'm looking for some kind of income or utility maximizing reason, not a clever way to punish red states for not having voted for President Trump.

Philo writes:

Fine, but exactly who is interested in practicing "the art of politics"?

TMC writes:

Scott, you put "dementia tax" in quotes, rightfully so as paying for your own care is not a tax. The problem with making healthcare a right is that it will then be a tax, and treated as such.

Scott Sumner writes:

John, Good example.

Andrew, I agree that the trend would be upwards in any case, but the subsidies make the level higher than otherwise.

Thaomas, Good idea.

On your second question, subsidizing spending in high tax states will tend to lead to higher than optimal spending, as residents of other states will bear part of the cost.

Philo, Politicians?

TMC, It's what economists call an implicit tax. That is, it's not literally a tax, but affects behavior in the way a tax would.

Capt. J Parker writes:

The art of politics is sort of like the art of boiling a frog. The key is to introduce sensible reforms so gradually that the public doesn't even know what hit them. Then a few decades later the public wakes up in an entirely different world.

What a repugnant thought. I thought this was the Library of Economics and Liberty. And such slim justification for handing over rule to the technocrats. If voters oppose eliminating a distorting tax break just maybe it's because they see a tax system where countless special interests have their own monogrammed carve out and they see no reason why their own ox should get gored if there isn't comprehensive reform.

And what's to ensure the frog boiling will be used for good instead of evil?

Michael Hubbard writes:

The Medicare entitlement and Obamacare are relevant policy changes that occur during the recent jumps in the last graph, right? Any others?

Weir writes:

If the reforms are enacted by democratically elected politicians, then they are democratic reforms.

But there's no legitimacy in getting elected. It's by relinquishing power, when you accept that you've been democratically un-elected, that you demonstrate your commitment to democracy.

That's what makes a regime democratic, the possibility of the rulers not ruling in perpetuity.

If instead you mount a rear-guard action to avoid having to give up your power, to avoid being democratically fired, then you're really not a democrat, and you ought to stop using that word.

This is a lot like the distinction between verification and falsification. It's also like, for example, when the EU was very disappointed with the Irish. They made them keep voting until they got it right. Even better, as Valery Giscard d'Estaing said, "is to avoid having referendums" to begin with.

And Mahmoud Abbas is in the twelfth year of a four year term.

Democracy is different because there's this possibility, from time to time, of changing course. It has nothing to do with any mythical public opinion. We're just talking about the rules of the game, and whether our betters are willing to acknowledge them, which would involve acknowledging some limit or constraint to their permanent rule.

You could put your faith in the administrative state, or the regulatory state, or the technocratic state, with its correct views on everything. But if there is no alternative, there's no means of ever correcting the blunders of our governments, which somehow or other do still happen, as bright and wise as the authorities are.

The important thing is, as ever, not to admit you were wrong in any fundamental way. Isn't that the first rule, above anything else? Even when it's entirely the government's doing.

For the mistake is Theresa May's. She's in charge. She wasn't "forced" to do anything. Yet somehow the elitists will always find a way to blame the populists for what the government's done. The public had nothing to do with it. Why blame them for what their rulers do wrong?

Scott Sumner writes:

Captain, If eliminating the tax deduction for health care is a bad idea, then don't do it. But I have yet to see any good arguments in favor of the deduction. And two wrongs don't make a right.

As far as "public opinion", I don't believe it exists. In a democracy, voters elect leaders that (we hope) will enact policies that make the country a better place. We don't have rule by public opinion polls, so I'm not sure what point you are trying to make.

Michael, Enacting Medicare and Medicaid almost certainly pushed costs higher.

Weir, Good comment. I agree.

Philo writes:

I suggest you reread you description of "the art of politics." Only a purely public-spirited politician would even try to practice it. Do you think our politicians are purely public-spirited? I thought George Masonites were devotes of Public Choice.

Capt. J Parker writes:

Dr. Sumner,
My point is that:

Governments are instituted among Men, deriving their just powers from the consent of the governed

I agree that there is no such thing as "what the people want." We all want different things. But, legislation enacted in such a way that "the public doesn't even know what hit them." negates any ability for the governed to give consent or even assent.

Brian writes:

In your post you seem to link your opinion on the non-existence of public opinion with the observation of framing effects. But those biases exist on the individual level also, so large-scale biases would appear to be more consistent with public opinion than not. Also, I think whether one accepts or rejects the existence of public opinion is largely a matter of taste. What is undeniable is that large bodies often acts in ways that are indistinguishable from single individuals.


Return to top