Economics/Psychology of Taxation

Since the recent SCOTUS health care decision, there’s been a lot of misinformation about taxation bouncing around the boards. Best to deal with it all at once in a new thread.

We start with a quiz, which I’m borrowing from the Nobel Memorial Prize winning economist Thomas Schelling. Two questions. I encourage everyone to try their hand at the questions, because they make an important demonstration. It would be super helpful, too, if you could record both your immediate emotional response, and also your more considered opinion after thinking about it for a while.

The topic is child exemptions in the tax code. A standard exemption is allowed for each child, up to two in this example. The amount of the exemption is independent of income. The first question is: Should the exemption be larger for the rich than the poor?

The majority of Schelling’s students at the Kennedy school at Harvard answered in this way:They found it completely unacceptable to give a larger exemption for the rich.Let us remember, though, that tax law is arbitrary. Above we had a childless family as the default, and we provided an exemption for having a child. Let’s change the default. Now a family with two children is the baseline, and families with less than two children will pay a surcharge on their taxes. The second question is: Should the childless poor pay as large a surcharge as the childless rich?

Please note your immediate reaction, and then give a more considered reaction. Schelling’s students responded in this way:His students rejected this proposal just as strongly as they had rejected the first proposal.I’ll give my own responses in the next post.

Here’s the answers.

Emotionally, my own responses were identical to the students. I wanted to reject both proposals. The only reason – and I mean that emphatically, the only reason – that I avoided giving the same answer “no” to both questions was my own economic experience. I’m about monetary policy, not a tax guy at all, but I knew enough about taxation to sniff out the trap. My initial emotional response to benefit the poor in both cases was, just as I had suspected, completely unsupportable. It is contradictory.

The two questions are related. To answer “no” to the first question is equivalent to answering “yes” to the second. I literally had to get out pencil and paper to work through this, or I would absolutely have gotten it wrong.

As Schelling notes,* the baseline that we use is completely arbitrary. The default position is arbitrary. It doesn’t matter. We can start with a baseline of no children, and then provide exemptions to people with children. Or we can change the default in our description of the situation. We can make the default two children, and then give penalties to people who don’t have the default number of children. These two different situations are identical. They are the same. The only difference is superficial, in the description, and this is what psychologists call framing. It’s like describing the freezing point of water as 32 Fahrenheit, or 0 Celsius. If you say that the temperature is too warm at 80 Fahrenheit, but too cold at 27 Celsius, then you haven’t fully thought through the implications of what you’re saying.

Both systems are precise. Both systems are accurate. But if you give different answers when you transfer from one formulation to the other, then you don’t yet fully grasp the underlying principle. Moreover, if we are intellectual unable to transfer freely between the equivalent formulations, then our thinking lacks precision entirely. In at least two other threads right now, there are posters who are actually rejecting, as meaningless or imprecise, formulations that are perfectly precise and indeed equivalent in meaning. They are, naturally, rejecting the framing that doesn’t fit with their own psychological pre-conceptions. They are only willing to accept the framing that puts their own position in a better light. They’re going with their emotions, at least so far, and thus have been unable to comprehend the very same situation framed in a different way.

If we cannot see the equivalence of these descriptions, then we are not yet fully qualified to talk about tax policy. It is as simple as that.

However, even saying that, I have no informed opinions about the constitutionality, or un-, of taxes and other stuff.
*I don’t have Schelling’s book in front of me, but I do have the excellent Thinking, Fast and Slow by the psychologist Daniel Kahneman, also a winner of the Nobel Memorial Prize in economics, who borrows Schelling’s example. This is a flippin great book, and I give it my highest recommendation to absolutely anyone not familiar with Kahneman’s work in psychology and behavioral economics.

Heh. This is a pretty cool demonstration of what we’ve been talking about. Thanks for providing it! I’m certainly no advanced mathematician, but I’m pretty good at the simple computation stuff, and as such, it seems bleedin’ obvious to me that the two positions are the same, to the degree that as soon as I saw there would be a second question, I knew what the second question would be. And I STILL answered from emotion the way the students did, although from intellect I answered that the quantities should be the same in either case.

While I can appreciate the thought experiment, two things jump out at me. One, the default of zero kids is not arbitrary. You are not born with two kids. If the goal of the policy is to incentivize having kids, you can achieve that result without “technically” burdening those who cannot have kids, or do not want kids. Nature set that default, not the IRS.

Second, as far a revenue is concerned, the two proposals have very different outcomes. Say I have a balanced budget, and we are about to vote which policy to enact. The first scenario means my revenue decreases as I will be handing back revenue I would have otherwise collected. In the latter scenario, my revenue increases, as I will now be imposing a fine on those who don’t have two kids.

Even though a surcharge and a credit are the same in many respects, they are not the same cognitively and emotionally, especially when the scenario is not zero-sum. Yes, the framing of the issue will affect the way people react, but this example is not particularly highlighting lazy thinking or dissonance IMO.

The law is an abstraction.

It can be written in any way.

“Nature’s default” does not have to relate, at all, to the logical default of legislation as it’s described. This is an important point because politicians are most emphatically not going to choose nature’s default when they’re selling a law unless it benefits them to do so. They’re going to choose the framing that provides them the most support for what they want to do. A mortgage interest deduction has a good chance of passing, as is demonstrated by the fact that it exists. A renter’s penalty doesn’t have a good chance of passing. That’s true for reasons entirely separate from what’s “natural” in the world.

This is an important point that psychologists like Keith Stanovich emphasize. If we cannot see the equivalence, then our decisions are made not by us, but by the people who manage to frame the issues for us.

This is a great example. Changes from the status quo are important.

This is absolutely right that changes from the status quo will affect people in ways that legitimately alter their opinions. True. Happens all the time.

And it’s still not legitimate to deny the meaning or precision of different equivalent frames. If we dislike the renter’s penalty, we can say that the renter’s penalty should be altered by making mortgage borrowers pay more. Or we can say that renters should pay less. Or hell, we can say that it’s okay for renters to pay a penalty. These are all potentially valid arguments. Lots of places where we can go from here.

But that doesn’t change the fact that there are equivalent ways to describe the current situation. The status quo is what it is, regardless of how we’d like to fix it. Sometimes, we’re just talking about what’s going on right now. We’re talking about a single situation. If people have different emotional reactions to a single situation, based merely on how the choice has been specially packaged for them, then that absolutely is lazy irrational thinking. What’s not valid is denying that the renter’s penalty exists, simply because you don’t like that frame. It does exist. It is a valid way of describing things. If we lack the awareness to see that equivalence, then that is sloppy thinking on our parts. That’s true, regardless of what we think the solution should be, or even if we think it doesn’t require a solution since it’s fine already.

No, it’s not. It’s a concrete reality. That’s why I think this thought experiment fails when it’s applied to a real-world scenario. Yes, the law is “made up” by people, but that does not mean every policy is some abstract notion based on whim and fancy. They are rules by which we govern ourselves that typically arise from an attempt to solve a problem, or achieve a specific goal via incentives and/or disincentives.

It can, but it’s not. That is a really important real-world distinction.

You are making two different points. Framing is very effective as a political tool. Of course any advocate will do their best to frame an issue in a way most advantageous to them. That said, a scenario that ignores a natural default is far less compelling than one that is truly random. People start with zero kids. Presumably, a policy which attempts to provide a mild incentive to engage in a very costly action (eg. having kids), is going to have very different real world consequences than a disincentive for not having kids. Doubly so in a situation where some people can never avoid this penalty through no fault of their own. This thinking makes sense if the tax is zero sum, meaning that the credit is truly a surcharge on someone else. Most often, it cannot honestly be framed that way.

Again, in theory this makes sense. In practice it doesn’t. Why? First, because the idea is to get incetivize home ownership. They do that by mitigating the substantial costs of doing something the government approves of (like buying a house). It’s kinda like an enzyme that speeds the process up by lowering the activation energy needed (money). A renter’s penalty raises the activation energy, if for no other reason than accumulating that energy is tougher since you will be subject to a tax the entire time you are attempting to meet that threshold. This is why tax credits and surcharges are functionally different even if they are logically similar.

Furthermore, the thought experiment creates a false dichotomy between two seemingly mutually exclusive groups. Plenty of people do neither or both. Also, in an efficient market, those mortgage interest deductions are passed on to renters in the form of lower rent. So helping owners does not necessarily harm renters. The same cannot be said for a renter’s penalty.

True, but framing is like 10th on the list of ways people mislead others. It’s much easier to lie, cheat, and steal, than it is to honestly frame something in personally advantageous way. This is why these things often don’t hold up in reality. Yeah, in theory, a doctor telling you you have a 75% chance of living sounds much better than you have a 25% chance of dying. But this ignores the fact that weighty decisions are rarely made solely on the basis of how an issue is framed.

But they are not equivalent in reality for some of the reasons I have already stated. There’s a reason there’s a carrot and a stick, not a carrot-stick. They both are used to achieve similar outcomes, but they are demonstrably different.

My first reaction was that they were asking the same thing. But then, I’m good at math, and my first reaction to these types of things is to look at them mathematically.

I am not good at math but it seems to me that the demonstration is flawed. Perhaps I need someone to explain it to me because it seems to me that there is no contradiction between responding negatively to both proposals. In the first I answer no because I believe the poor should pay a lower effective tax rate than the rich. To the 2nd I also answer no because again I believe that the poor should pay a lower effective tax rate than the rich. I guess the economist believes that by answering “no” to the first question a person is agreeing that deductions should be equal. This is a false assumption. Am I missing the point completely?

This is how I looked at it, the poor should NEVER pay as much or more tax than the rich period. So I feel like I am missing the clever gotcha bit?:dubious:

Let’s break it down.First Baseline of Tax Burden, No Children
Poor Man: 1,000
Rich Woman: 20,000Poor Man has a very low income, Rich Woman high. Poor Man doesn’t pay much in taxes. Rich Woman pays a lot more. You can see how much they give to the gummint, 1k vs 20k. We as a society wish to encourage the sexings, so we want to provide some incentive for baby making. A child tax “exemption” is the proposed solution. First question from the quiz: Should the exemption be larger for the rich than the poor? We answer hell no. In fact, to make the math easier, let’s be even more blindly generous, and say that Poor Man should get a bigger exemption than Rich Woman. It’s only fair, right? Let’s give him a $500 exemption, and her just $200. She doesn’t need it as much.First Baseline of Tax Burden, Minus Exemptions for Kiddies.
Poor Man: 1,000 - 500 = 500
Rich Woman: 20,000 - 200 = 19,800There we go. Poor Man gets a bigger exemption for breeding. Good job, guy. Way to pay less tax for spreading some seed. Ah. But remember that the baseline for taxation is utterly arbitrary. There is no right answer here, no perfect baseline, just like there’s no correct choice between Fahrenheit or Celsius. So now let’s set the baseline tax burden as two children. This is our updated baseline, our new default level.Updated Baseline of Tax Burden, With Children
Poor Man: 500
Rich Woman: 19,800I’m not intending to create a new tax system here. I want an identical tax system. I want the amount that people pay, the final calculation, to be exactly identical to the last system. The only thing that’s different here, the only thing, is that I’m describing the default differently. Now the default, the new baseline, is a family with children. This tax system is the same, so people with children pay less tax. But because the baseline is different, I need to calculate what the penalty is for not having children. This “penalty” is what it absolutely logically must be, mathematically, to match the first situation. We’re not changing the tax system, just changing the way we describe it. Let’s add it up.Updated Baseline of Tax Burden, With No Kids Penalty
Poor Man: 500 + 500 = 1,000
Rich Woman: 19,800 + 200 = 20,000And lookit that. The numbers for no children are exactly the same as they were before. I can’t emphasize that enough: This is the same tax system. There are no changes in the tax system, except the terminology. We’ve switched from Fahrenheit to Celsius, and that is all. People in both systems pay the exact same amount of tax for any given situation. Water still freezes as you would expect. There is no difference here between the systems. Yet you can see clearly that Poor Man has a bigger tax “penalty” for having no children. Why? Because providing a bigger exemption for a certain behavior is absolutely the same as providing a bigger penalty for not doing that behavior, when you consider the relative baselines.

There is no such thing as having a bigger tax credit for one behavior, without having a bigger penalty by not doing that behavior. If you want to give Poor Man a bigger exemption for having rugrats running around, then you are by mathematical/logical/economic/rational necessity giving Poor Man a bigger penalty by not having rugrats running around. There are not two different concepts here. There is one concept here, that often goes by different names depending on who prefers which policy. Both tax systems are identical. They are the same. Poor Man with no kids pays the same amount in both systems. Poor Man with kids pays the same amount in both systems. The systems are the same. They are utterly identical.

The only difference is the framing. This really drives some people through a loop, but it’s true.

People who like certain exemptions will of course twist their language to talk about what’s honest or natural or whatever. Sheer empty emotionalism. The calculations don’t change. The two systems are not actually two. There is one tax system here, only one, with two different ways of describing it. Which description you prefer is going to depend, entirely, on whether your personal ox is getting gored. My recommendation, if we want to talk about this stuff honestly and thoroughly, is to get very very accustomed to switching back and forth. If we can’t do that, our emotional attachment to one description or the other is going to get the better of us, and then we’re going to start saying damn fool things like that the description we subjectively prefer is somehow more “precise” or “honest”.

Our emotions are too strong on subjects like this. We must be able to change our mental framing at will to have clarity of thought.

A car is an object of steel, plastic, glass, etc.

The law is a set of beliefs set in words, ink on paper in a bizarre interconnected pattern which only has meaning in our minds.

One of these is concrete reality.

The other is quite obviously not.

When our abstract words have multiple ways of describing the same situation, a change in the words need not, by itself, change the referent, the situation being described. That’s one of the big deals with symbolic abstraction, of which the law is one of the most notable examples in all of human history, maybe second only to the broader notion of written language itself. This abstraction provides many possibilities. A car is just a car. It is concrete. It is what it is. But abstractly, it’s possible to define two objects, for example a set for which all open covers have a finite subcover, and a set which is both closed and bounded, and despite the fact that I have seemingly defined two different things, it turns out that by following the basic rules of logic/math, we end up seeing that we’ve defined the same thing, in this case a compact set. The tax example above is like that. It looks like I’ve defined two different tax systems. Untrue. There is one tax system, only one, described in two different equivalent ways. The car is what it is. The tax system, however, is at first seemingly two different things, and only after following our abstract rules of logical implication can see that we don’t have two different systems here. We have the same system, defined with different words.

With pure math there’s no psychological interference. With applied math, we often need to overcome that bias to see the underlying equivalence of structure. If we can overcome what’s psychologically compelling, we can see they’re equivalent. Maybe you think this isn’t important, but I would strongly disagree. I would say it’s a necessity.

If we can’t get simple questions right – Fahrenheit vs Celsius, subsidy vs penalty, abstract vs concrete – then we have little hope of consistently getting bigger questions right. Which brings me to:

This is, of course, wrong.

First, it’s one of the foundation principles of public finance, so important that it’s often taught in freshman intro to econ classes, that the legal incidence of taxation is irrelevant to the economic incidence. In other words, it doesn’t matter who you tax: the final result is the same. Unless you have amazingly inelastic demand for rental housing, there is no possible way for the deductions that benefit landlord mortgage borrowers to accrue to their renters at the same rate that personal mortgage borrowers directly gain that benefit. It is, in other words, inevitable that renters will be harmed relative to borrowers. That is necessarily true.

But more than that, I’m not talking about a shift in the status quo. I am not talking about a newly implemented mortgage deduction (which can be equivalently described as a newly implemented default rate, plus a renter’s penalty). I am talking, as I have been from the beginning, about an analysis of a single tax system. A relative benefit for one group can be equivalently described as relative burden on the other. It’s always going to be so.

This thread has never been about two different tax systems. It’s not even been about changing the tax system to see who benefits. From the OP, I have been talking about one tax system, just one, which can be described in various ways. I have been emphasizing the equivalence of the two descriptions. They look different, but they’re not. The calculations are the same. The tax burdens are the same. The amount of money that goes to the government is the same. The only difference is the description. So on top of the note about economic incidence, I want to emphasize that it is utterly impossible for a mortgage interest deduction, and its equivalent renter’s penalty, to have different economic effects. They don’t have different effects. They cannot possibly have different effects. They have the same effect, because they are the exact same calculations, described in different words. There is no difference between them. They are the same. I can run a similar example like I did with the baby deduction to demonstrate that.

That is the whole point of this. A difference in words does not necessarily convey a difference in underlying structure. It’s important to know this when some words are emotionally distracting.

If we start with a baseline of a higher tax rate, and institute a mortgage interest deduction, that’s one way of describing things. Even if we grant that landlords in addition to personal homeowners can take advantage of it, and pass on some of the benefit (absolutely not all) to their renters, I can still describe an equivalent system by adjusting the arbitrary baseline and instituting a renter’s penalty. In that case, the default rate would be low and there would be a penalty for renting, or at least for not borrowing money. Landlords who qualify for the low baseline rate (identical to the previous deduction) would be able to take advantage of that lower rate, and that relative low rate would then be passed partly to renters, in the exact same measure that it was passed on before. Renters would then pay a relative penalty, and it would end up being a penalty that led them to pay the same burden that they paid originally under the higher baseline.

Equivalent. All of these descriptions of deductions and penalties can be reframed in equivalent ways.

I’m not proposing we go through the entire tax code, rewriting it in different words (although that theoretically could be done, if they ever stopped changing it). I’m saying we should realize that every penalty of this sort can be seen equivalently as a deduction. Same system, different descriptions. The terminology we choose doesn’t affect the calculations. Even shifts from the status quo, if I eventually wanted to talk about that, could be described in equivalent ways which superficially seem vastly different.

Being able to do this stuff before getting into the nitty gritty of taxes is, I think, extremely important.

I’d argue it has broader benefits, too.

That’s the gotcha. The wording is set up to make it seem like the first is hurting the rich while the second is helping the poor. But both are the same.

When you have two groups within society you can either offer one group a deduction or impose a surcharge on the other.

The mortgage tax deduction is a little more straight forward. People either rent or own. If we give the people that own a deduction, the net result is that people who rent pay a surcharge. Since at the end of the day, the people that rent will pay a higher effective rate than the people that own.

What this shows is both how important the message is, but also how stupid people are when it comes to taxes that they are so easily manipulated. So many good hearted people start with the idea “the poor should NEVER pay as much or more tax than the rich period” and will vote based on how the message is manipulated towards them.

Death tax.
Marriage penalty.
Health tax.

People aren’t able to sift past the message to see how deductions and surcharges end up being the same thing.

I admit, I only got this right because the way you posed the question primed my brain, so I recognized that the example looks very faintly like the intuition of the Coase theorem. I went in expecting the scenarios to be equivalent. But it definitely takes some work to put the normative baggage aside. I’m certainly as prone to heuristics and biases as anyone else.

Again, this is only true if you rig the system to get the same outcome in both examples. In the real world, it usually doesn’t work that way.

I disagree. People respond to incentives and disincentives very differently. That alone, for a system run and occupied by people means that the two systems are not the same.

The “law” is technically words on paper, but the force of law, which is what makes the law the law, is overseen by people with guns, tanks, and jail cells. That’s why just saying it’s an abstraction is misguided. Violating the law will in most circumstances introduce one to the real and concrete nature of the law which has all the realness of a car or another physical object.

We have, in fact, two different systems because the total outcome of those system (eg. the aggregate revenue) will be different. Despite your rhetoric, in the real world, a tax an subsidy are different for a number of reasons.

I will give you an example. Say I own a store. I decide in order to stimulate demand, I will offer a Groupon. Let’s for simplicity’s sake say the I issue 1000 coupons for 25% off. Now, let’s say I have a 3000 regular customers, 1000 of whom will buy the groupon. By your logic, I have imposed a price penalty on the other 2000. The problem with that is that the price they pay has not changed. It has only “changed” relative to what others will got the coupon will pay. Now, you get around that by assuming that I will change my prices in a way to make my coupons revenue neutral in aggregate. The problem is that is generally not how the world works. I am usually not worse off for you having something I don’t in a non-zero sum situation.

What distinction are you making between pure math and applied math?

It’s not a relative comparison. Let’s say you the interest deduction disappeared tomorrow. What would happen to rents? I happen to rent a condo. If my costs go up, I will pass it on to my renter. If my costs are subsidized along with most other owners, the price people will pay is likely going to be lower.

That’s where you err. There is always a status quo. You have to start from something. Even a newly implemented policy is rests upon, and is subject to influence from other policies. A newly implemented mortgage deduction rests upon a system of taxation that allows deductions, etc., etc.

The money that goes to the government is only the same because you anchored yourself to a number in both examples. I can make a flat tax, a value added tax, and progressive income tax system, all bring in the same amount of revenue if I change the numbers haphazardly, but it doesn’t mean each of the systems is the same, or that it will have the same outcomes, or that it will incentivize the same behaviors.

But you are forgetting we will always start with the initial baseline. I the first example you do one thing: give a subsidy. In the second you have to do two things: adjust the baseline and institute a tax. You justify this by saying the first baseline was arbitrary, so changing it fine, and that as long as we change it in a way to make the outcomes the same, nobody should care. The problem is that making the outcomes the same is more difficult to achieve than you imagine, and that we will never have a world full of coldly, rational people. Yes, this can often be a bad thing, but it is often not an issue.

One more point on this. You seem to be implying that “pure math” is better even when applied to real world circumstances, so I would like to pose a few hypos to you.

I have three game whose odds are all based on the flip of a coin. Here are the three different scenarios:

  1. Heads you win $5, tails you lose $4
  2. Heads you win $50,005, tails you lose $50,004
  3. Heads you win $500,005, tails you lose $500,004

Question. Are the games all the same? Say you get to play for 1 hour. Would you feel as comfortable paying game #3 as #1. What does the pure math tell you is best?

Hypo number two. Say we have another game with coin flips. Heads you win $2, tails you lose $1. You examine the coin and it seems true. Yet, after watching this guy for two hours, you have yet to see him flip heads. He asks you if you want to play. Do you play? What does the math tell you to do?

Do you think the motives of someone trying to punish someone for not doing something are the same as someone trying to reward someone for doing it?

The example only works because we set this up as an unindexed deduction from taxable income. We could, for example, in the 2nd case set it up a tax credit that is dependent on your income, phased out as your income gets higher. In fact, we may even do that to some extent in the US tax code now, except that you have to have a pretty high income to hit the phase out point. I know that some deductions are phased out like that, but not sure about the child deduction.

There isn;t a “child deduction”, it is an exemption. In recent history (up to 2009, I think), both deductions and exemptions have been reduced above a certain income level.

On reading the OP, I thought the thread was going to be about the value of exemptions to the rich versus the poor. An exemption for dependents means zero or little to poor people, who have either zero taxable income, or are taxed at low amounts, such as 10%. Conversely, high income earners benefit from saving up to 35% from exemptions and deductions.

Because of this, the OP is flawed. The two cases are not the same. In the first case, the poor family gets an exemption worth zero (or close to), yet in the second case is made to pay a tax surcharge. The second case costs more than the first case saves, and hence the two scenarios are not equivalent.

When explaining this to math-challenged people in the future I would lead with this. The breaking it down didn’t help me grasp it at all and it wasn’t until I got this far that it became clear to me.

First off, we Libertarians would say, what does my choice of habitation have to do with financing the government? There should not be a deduction for anyone. We all have to live somewhere! But, I’m sure your government knows what is best for us.

BUT, given the scenario, consider:
2 single childless people working at the same place making the same income. If their taxable income is the same, then their tax burden (obligation to finance the legitimate operation of the government) should be the same. However, if one rents and one buys, the costs of each being the same amount, it may be said that the buyer is relieved by a certain portion by being able to deduct the interest on his mortgage making his taxable income less than the renter, OR, it may be equally said that the renter is punished by having to contribute more toward financing the government than the buyer.

However, all things are not equal. If it were figured up so the tax rate were reduced to effect a neutrality for renters and buyer, but then added that renters must pay an extra X amount per year penalty, the number of people renting would probably go down.
It doesn’t matter if you hit me with the stick and I run 5 mph or you hold out the carrot and I run 5 mph, I am still playing the hamster.

Same could be said for little people:
Same 2 people working at the same and all the same, except one decides to produce a little tax exemption to get rid of his disposable income and one decides he would much rather buy a big old boat to throw away his disposable income. For some strange reason the government has decided that boat owners must pay more in taxes than the makers of noise makers. Just looking at the same situation from different sides.

I predict this will go badly. The local cool kids will be along shortly to bully you. You might want to take a few minutes to prepare arguments against things you’ve never believed.

It’s true “Libertarian” covers a very wide variety of viewpoints, so it might be helpful for him to be more specific.