{"items": [{"author": "Sean", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654769449652", "anchor": "fb-654769449652", "service": "fb", "text": "Taxes do not deter behavior. A Tobacco Tax deters people from smoking like the Income Tax deters people from working. We should limit an activity where a person's individual rights end and another's begin. For instance in Concord they passed a bottle water ban to deter a behavior and support a clean environment. However all they have done is stop people from buying a product that is good for them and 100% recyclable, and left products full of sugar and twice the amount of plastic and harder to recycle. No behavior was deterred, the only thing that was stopped was a healthy and better option. In my opinion the idea of banning or taxing something to stop behavior that is stoping people from being independent", "timestamp": "1397101688"}, {"author": "David", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654771490562", "anchor": "fb-654771490562", "service": "fb", "text": "\"A Tobacco Tax deters people from smoking like the Income Tax deters people from working.\"
Which is to say, it deters a bit? Most empirical estimates I've seen find an elasticity of tobacco around -0.4 (i.e., a 1% increase in price reduces tobacco consumption by 0.4%). Estimates of the elasticity of income with respect to the net-of-tax rate is are usually around -0.25 to -0.4 for people on average incomes.", "timestamp": "1397102647"}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654771939662", "anchor": "fb-654771939662", "service": "fb", "text": "Well the income tax can also incentivize working in a way I believe to be unlike the tobacco tax, which I don't think has any plausible way of incentivizing tobacco consumption.", "timestamp": "1397103105"}, {"author": "Sean", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654774025482", "anchor": "fb-654774025482", "service": "fb", "text": "David, in Boston alone 40% of cigarettes are black market... for all the \"benefits\" this tax creates it causes a boom in profits for smugglers. The war on tobacco is creating a new form of organized crime and increases will only lead to a drop in revenue. And isn't it hypocritical for the state relying on revenue from tobacco users while at the same time saying it doesn't want people to use tobacco??", "timestamp": "1397105154"}, {"author": "David", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654774729072", "anchor": "fb-654774729072", "service": "fb", "text": "I agree that taxes on cigarettes shouldn't be raised too high because people will shift to black markets; I don't know what the optimal level is and I certainly can't speak for Boston's experience, but it is nevertheless the case that some level of tax on cigarettes will reduce smoking. And this is all we need for it to be relevant to Jeff's original post.

\"And isn't it hypocritical for the state relying on revenue from tobacco users while at the same time saying it doesn't want people to use tobacco??\"
I don't see this as being hypocritical at all. If cigarette consumption (and the government's tax take from it) falls, then I'm sure the government will be able to tweak some other tax to make up the lost revenue if it needs to. It would be hypocrisy if the government _encouraged_ people to smoke while taxing it ostensibly because smoking's bad for your health.", "timestamp": "1397105925"}, {"author": "Adam", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654782249002", "anchor": "fb-654782249002", "service": "fb", "text": "One obvious case where taxes are a very bad idea, when the government is attacking the same problem in several ways.

Governments want to increase tax take, hence they want to encourage (c.p.) any taxed activity. this can cause policy conflicts. (It is often claimed that this is one of the reasons anti-smoking laws are less stringent than they might be).", "timestamp": "1397122617"}, {"author": "Adam", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654782273952", "anchor": "fb-654782273952", "service": "fb", "text": "George: \"Well the income tax can also incentivize working\". What way is that?", "timestamp": "1397122678"}, {"author": "Andrew", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654782288922", "anchor": "fb-654782288922", "service": "fb", "text": "There are cases where it is easier to enforce a tax. For example in Houston, Texas, people will sneak out at night to water their lawns.", "timestamp": "1397122853"}, {"author": "David German", "source_link": "https://plus.google.com/111229345142780712481", "anchor": "gp-1397133688268", "service": "gp", "text": "As a rule of thumb, I'd suggest taxing things that have externalities, and banning things that have victims.\n
\n
\nI don't think either is applicable to the lawn-watering example. \u00a0There might be a small environmental externality, but most of the harm is in diverting water from other uses. \u00a0Market pricing alone should take care of it - and if lawn owners in California are more willing to pay than rice farmers or cattle ranchers, what's wrong with that?", "timestamp": 1397133688}, {"author": "Hassan", "source_link": "https://plus.google.com/109686970531250960199", "anchor": "gp-1397136676197", "service": "gp", "text": "A tax isn't always a useful tool if the purpose of a product is varied and relatively opaque from a usage perspective. We can tax tobacco at whatever rate we find effective, because it has a narrow, non-essential use. This isn't true for something like water - we don't want people wasting it on their lawn, but we certainly want it to be available as cheap as possible for people to drink, wash, etc.\u00a0\n
\n
\nThe obvious answer is to graduate taxes by usage, so that people pay more per unit the more they use. But this is also imperfect - a household with a lot of people in it might find themselves running into a harsh tax, while a single person could water their lawn without a penalty, which isn't what we want.\u00a0\n
\n
\nIt seems like there are enough places that have banned lawn-watering during droughts that we should be able to determine whether it is actually effective in lowering people's usage. Certainly some people will cheat, but that might acceptable if it generally suppresses use without punishing legitimate needs.\u00a0", "timestamp": 1397136676}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1397144950834", "service": "gp", "text": "Wealth-signaling in general seems like an excellent situation for taxing over banning. If lawn/car/pool sizes are for status, then it's their relative size that matters. With high taxes, the status-competitors can't afford such large things, so they compete with smaller ones. Government gets money, externalities are internalized, and no one is really worse off. (This is also an argument for taxing such things even in the absence of externalities.)", "timestamp": 1397144950}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1397145024990", "service": "gp", "text": "Your lawns example is a little weird b/c of the variability in water availability and difficulty in changing the size of your lawn. But I'm not sure those facts are so important even in that case.", "timestamp": 1397145024}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1397145144944", "service": "gp", "text": "@David German\n\u00a0-- I'm not sure how clear the distinction between victims and externalities is.", "timestamp": 1397145144}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654838910452", "anchor": "fb-654838910452", "service": "fb", "text": "To be clear, the actual incentive effects can probably go either way. We should just maximize the Laffer curve empirically and I suspect that would involve much higher top marginal rates.", "timestamp": "1397163745"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654847852532", "anchor": "fb-654847852532", "service": "fb", "text": "George: The peak of the Laffer curve is not the utility maximizing point. It is the short-run taxation maximizing point, but as private consumption and investment have positive utility, and are maximized to the left of the Laffer peak, utility is maximized at a tax rate strictly less than that which would maximize taxation.", "timestamp": "1397169349"}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654854549112", "anchor": "fb-654854549112", "service": "fb", "text": "Raising revenue for government programs also has utility.", "timestamp": "1397173525"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654855552102", "anchor": "fb-654855552102", "service": "fb", "text": "That is entirely consistent with what I just wrote. If tax revenue yields positive utility, the utility maximising tax rate will reside between the private sector maximising tax rate and the tax revenue maximising point - to the left of the peak of the laffer curve.", "timestamp": "1397174048"}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654855786632", "anchor": "fb-654855786632", "service": "fb", "text": "I guess what I am saying is that with my utility function I want something close to the peak since I want to maximize redistribution with progressive taxation and raise a lot of revenue.", "timestamp": "1397174193"}, {"author": "Arthur", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654875157812", "anchor": "fb-654875157812", "service": "fb", "text": "Ben, if you believe that government spending and investment has greater utility than private spending and investment as a rule, then the utility maximizing and revenue maximizing peaks are the same.", "timestamp": "1397184540"}, {"author": "Daniel", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654886075932", "anchor": "fb-654886075932", "service": "fb", "text": "Arthur, not necessarily (and probably not at all, IMO). Unless you believe private spending has a net utility of zero or negative, that statement isn't logically sound. (I assume you don't believe that, even if you do believe its utility is less than that of government spending.)

There might be a point near that revenue maximum where private spending decreases $4-5 for every dollar for every $1 of tax gain. Even if government spending had twice the utility of private spending, this case would mean that approaching this revenue-maximum peak would take you further from the utility-maximizing peak.

Though the numbers I chose may be more extreme than real-world numbers to make a point, I don't think the general situation is unrealistic. The reason we have a peak at all is because at a certain point taxes discourage people from working more than they increase revenue through the increased proportion of the income still produced. There is very likely a range, maybe a substantial range, where an increase of taxes will decrease work/production ALMOST (but not quite) as much as the increased proportion of income increases government revenue. You're still heading towards the revenue maximization peak by increasing taxes in that range, but it probably isn't worth it for the country as a whole, so not maximizing utility.", "timestamp": "1397188579"}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=654887667742", "anchor": "fb-654887667742", "service": "fb", "text": "Empirically maximizing the Laffer curve with some sort of flat tax is very different from doing it within a highly progressive tax code and just for the wealthy taxpayers, which is what I am interested in. If we restrict the empirical maximization to the ones we want to tax a lot anyway, those people probably don't need to spend much of their income (if they did they wouldn't be rich) and the example numbers Daniel just presented might be much less realistic.", "timestamp": "1397189223"}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1397242267336", "service": "gp", "text": "@David German\n\u00a0\"Market pricing alone should take care of it - and if lawn owners in California are more willing to pay than rice farmers or cattle ranchers, what's wrong with that?\"\n
\n
\nThe benefit of a lawn is mostly relative status, while the benefit of rice or cattle is enjoyment. \u00a0If we ban lawn watering people are mostly about as happy as before while if we ban rice-farming people are not.", "timestamp": 1397242267}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1397242913166", "service": "gp", "text": "@Hassan\n\u00a0\"we don't want people wasting it on their lawn, but we certainly want it to be available as cheap as possible for people to drink, wash, etc\"\n
\n
\nThe standard economics answer to this is \"set the price at the point where supply meets demand, and give poor people money\".", "timestamp": 1397242913}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1397243051124", "service": "gp", "text": "@Hassan\n\u00a0\"graduate taxes by usage, so that people pay more per unit the more they use\"\n
\n
\nI think this is about equivalent to \"charge a flat rate of $X/gallon and give everyone $Y\", with the advantage the under the \"give people money approach\" if someone would rather use almost no water and have more money we will let them.", "timestamp": 1397243051}, {"author": "David German", "source_link": "https://plus.google.com/111229345142780712481", "anchor": "gp-1397248523009", "service": "gp", "text": "\"The benefit of a lawn is mostly relative status\"\n
\n
\nCitation needed? Lawns are nice for all sorts of things. ", "timestamp": 1397248523}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1397249143186", "service": "gp", "text": "@David German\n\u00a0Yes, this argument only applies if it's true that lawns function mostly as positional goods, and perhaps they don't. \u00a0But if they do, or if we find other goods that really are mostly positional, what do you think?", "timestamp": 1397249143}, {"author": "David German", "source_link": "https://plus.google.com/111229345142780712481", "anchor": "gp-1397274401423", "service": "gp", "text": "I don't know how distinguish a \"positional good\" from a taste I just don't happen to share.\n
\n
\nI do think history shows that prohibiting victimless activity fosters black markets and violent gangs, invites official corruption and regulatory capture, and generally undermines the rule of law. \u00a0Ditto for sin taxes to a greater or lesser extent, depending on the magnitude of the tax.", "timestamp": 1397274401}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=655103315582", "anchor": "fb-655103315582", "service": "fb", "text": "Arthur George Here\u2019s a simple proof that the overall utility maximizing point is less than the Laffer maximum.

Notation:

P=Private Sector GDP

T=Taxation

Y=GDP=P+T

r=tax rate.

P,T,Y all positive real numbers. 0 <= r <= 1.

We can restrict the analysis to consider only the P and T corresponding to an arbitrary sub-population. For example, we might consider P=Private Sector belonging to \u2018the rich\u2019, T=taxation paid by \u2018the rich\u2019. This does not change anything with the maths.

Up(P)=utility generated by private sector consumption.

Ut(T)=utility generated by taxation.

Uy(P,T)=total utility

Note that Ut(.) might include utility generated by spending the tax raised on groups other than the population the taxes are being raised from. E.g. if we are restricting our attention to taxes raised on high earners, we still consider the utility generated by spending their taxation on others.

For simplicity, we assume Up() and Ut() are linear (It would be better to assume log-linear but again I am trying to keep this simple and it will be evident from the proof that the actual assumption we need is much weaker)

Up(P) := A*P , A>0

Ut(T) := B*T , B>0

Uy(P,T) := Up(P) + Ut(T)

Note that we make no assumptions about the relative sizes of A and B \u2013 one could be much bigger than the other and it would make no difference to the argument.

Let r0 be the tax rate such that dP/dr = 0. Note this may or may not be r0=0. For simplicity, we assume dP/dr < 0 for all r>r0

Let r1 be the tax rate such that dT/dr = 0, the laffer maximizing tax rate. As per the standard Laffer curve, we assume r1 < 1

As Y is finite, P = Y*(1-r) = y*0 = 0 at r=1. Hence r0 < 1.

We need a quick lemma:

For r < r0, 0<dP/dr = d( Y(1-r) )/dr = (1-r)*dY/dr + Y*d(1-r)/dr = (1-r)*dY/dr \u2013Y

As 1-r => 0, this shows dY/dr>0 for r<r0

dT/dr = d(r*Y)/dr = Y + rdY/dr > 0+0
hence r1 > r0

Now we simply note:
As r0 < r1,
at r1:
dUy/dr = A*dP/dr + B*dT/dr < A*0 + B*dT/dr = A*0 + B*0 = 0+0 = 0

Hence utility is decreasing at r=r1, the laffer maximizing tax rate

As r is in [0,1], a closed interval,
Uy is maximized by some r < r1.", "timestamp": "1397276617"}, {"author": "Daniel", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=655130560982", "anchor": "fb-655130560982", "service": "fb", "text": "Thanks, Ben. I haven't done much serious math like that since graduating college.

George, Republicans talk a lot about how raising the tax rates on the rich de-incentivizes productivity and is bad for the economy as a whole (even those who are not being heavily taxed.) Although I think their worry about this is wildly overblown, because the effect is minimal at/near the levels of taxation we actually have in the US, I do think it's a real thing that can happen if taxes get high enough. Yes, even in the highly progressive system you suggest, where us normal-income folks aren't overtaxed. We will never get there in this country, so I'm all for raising taxes on the rich relative to what we have now, but approaching that revenue-maximum too closely is really not a good idea in ANY tax bracket.", "timestamp": "1397311391"}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1397324582804", "service": "gp", "text": "The \"position good\" distinction seems real (defined like: \"is your happiness more a function of how much you have, or how much you have relative to others you're comparing with?\"), and while I don't know how to measure it exactly, I'll bet that's something economists have studied a fair amount. (E.g. probably something about the shapes of various supply/demand curves.)\n
\n
\nStill, I disagree with Jeff that positional goods are ripe for bans. As I said, if anything they'd be a great area for taxes (since everyone vying for the positional goods could have the same relative position and so the same happiness, but we have additional tax revenue).", "timestamp": 1397324582}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=655195815212", "anchor": "fb-655195815212", "service": "fb", "text": "Ben, your proof contributes nothing to the discussion as far as I can tell. I bet I could add a similarly worthless proof of the following statement: Suppose that 0 < r* < 1 and r* is the revenue maximizing tax rate. For any epsilon > 0, there exists A,B, with similar assumptions on the U's where r*-epsilon maximizes utility in your setup.

As one might guess from reading my earlier remark \"with my utility function I want something close to the peak,\" I don't really care if something optimal is a bit less than the theoretical peak of the Laffer curve (which we don't have access to anyway). I am not a utilitarian and I don't actually make decisions using assumptions compatible with your proof anyway. For example, my own personal societal utility function might have a delta function component centered on the Laffer peak. But it doesn't really matter since no one really cares about questions like \"given a bunch of assumptions, do we want a tax rate equal to some specific real number of less than that real number.\"", "timestamp": "1397350464"}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=655196503832", "anchor": "fb-655196503832", "service": "fb", "text": "When creating a tax code, there are two central questions. 1. How much revenue should we raise? 2. How should the tax burden fall across the population? My answer to the first one is probably something like 42% of GDP. My answer to the second one is that we should raise revenue in a highly progressive way that also discourages things like pollution/fossil fuel burning. So progressive income, consumption, and possibly wealth taxes, a carbon tax, and maybe a land value tax along with the necessary transfers to keep things highly progressive.

My whole reason to even bring up the Laffer curve is to say that we shouldn't be afraid of the work disincentive effects because we can measure them empirically and back off a bit when we start reducing revenue.", "timestamp": "1397350911"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=655196678482", "anchor": "fb-655196678482", "service": "fb", "text": "\"Arthur: Ben, if you believe that government spending and investment has greater utility than private spending and investment as a rule, then the utility maximizing and revenue maximizing peaks are the same.\"

I was showing this was incorrect.", "timestamp": "1397350978"}, {"author": "George", "source_link": "https://www.facebook.com/jefftk/posts/654766919722?comment_id=655196858122", "anchor": "fb-655196858122", "service": "fb", "text": "I would argue that you took Arthur's comment too literally and that he actually meant \"close to the Laffer peak.\"", "timestamp": "1397351049"}, {"author": "Marcus", "source_link": "https://plus.google.com/111675838261170541573", "anchor": "gp-1397580174670", "service": "gp", "text": "There's also an interesting distinction between a tax as a market mechanism, and a cap-and-trade: it turns out which is the better regulation depends on the form, volatility, and the relative uncertainty of the cost & damage functions. A damage function with a strong threshold makes a cap the right approach. If the damage function is more uncertain than the cost function, generally means a tax makes more sense (and vice versa).\u00a0 If the cost function is volatile year-to-year then a tax also makes sense. etc.", "timestamp": 1397580174}]}