Thursday, September 18, 2008

taxes

I cannot claim to be anything remotely close to an economist (though neither can McCain or Obama). Still, some (to some extent unconventional) very high-level thoughts about taxes - thoughts not having to do with complicated graphs, econometric data, etc. (and so high-level thoughts I have a hard time seeing refuted by complicated graphs, econometric data, etc., though I'm open to being corrected on that).

1. The level of taxation only affects WHO will be spending HOW much money (in particular, private individuals generally vs. private indivuals who get money from the government vs. the government itself), and so only affects WHICH sectors of the economy get stimulated. As a result, there is nothing intrinsic to leaving money in the hands of private individuals (so that they can spend it) such that doing so necessarily leads to more/better economic stimulation. And, at least when it passes through the hands of the government -- IF the government does a good job -- it can consciously and intentionally direct where money gets spent so that the economy is overall more/better stimulated.

2. The traditional, Adam-Smith-type, "invisible hand" argument for leaving private individuals alone to spend money how they wish has nothing to do with economic stimulation. Rather, it has to do with with efficiency in satisfaction of needs/desires on the part of goods and services produced by the economy, whatever its degree of "stimulation". And where (unlike old Soviet 5-year plans, which involved the government itself placing [very enormous] orders for goods and services) the government simply takes tax money and gives it to other private individuals (almost always with restrictions, of course, e.g., Medicare), it seems less like a question of WHETHER needs/desires successfully get met by goods and services, but rather WHOSE. (So the "redistributive taxation" of welfare capitalism is, in at least one very important respect, very different from the spending of Soviet-like centralized governments.) In other words, when it comes to the issue of "information transfer re. needs/wants across various sectors of the economy", there is nothing intrinsically inferior when it comes to welfare capitalism (as compared to pure free-market capitalism).

3. Another argument for leaving money in the hands of private individuals that I often hear is: when this happens, many private individuals will save some of that money, leading to a larger supply of money in the lending market, reducing interest rates, increasing entrepreneurship, etc., stimulating the economy. But (and here is in effect a very special case of point 1 above) absolutely nothing prevents the government, after acquiring money through taxation, from itself putting money into the savings market.

4. Taxation does inherently involve unavoidable money lost due to administration and transaction costs. However, that does absolutely nothing to show that more rather than less taxation inherently involves unavoidably greater money lost due to administration and transactions costs. E.g., however high or low taxes are, the IRS will get the same number of tax returns in the mail every year. If anything, all these exceptions and special cases ("loopholes") which help to reduce taxes for some (much of this "some" being very rich individuals) only add to administration and transaction costs.

5. A different argument I often hear is: taxation disincentivizes, i.e., people are less likely to work harder (in order to get that promotion, or whatever), since what they might possibly get from working harder, namely, higher income, will be compromised by money lost to taxes. Undeniably, taxation disincentivizes IN THIS RESPECT. But if, e.g., the government directs more money towards healthcare, there becomes a greater demand in various labor markets related to healthcare, increasing salaries with such jobs, incentivizing them. So there can be no general argument that taxation necessarily disincentivizes OVERALL. And, when the government is consciously and intentionally directing where large-scale amounts of money go, it can, in principle (at least if it knows what the hell it is doing), direct it in ways which do better by way of incentivizing overall. In fact, nothing at all stops the government from "reverse taxation" for some individuals, i.e., adding some (relatively) fixed percentage to their income, for purposes of incentivizing.

6. Another very common argument is empirical: examination of historical cases of tax increases and decreases shows an overall boost to the economy when taxes are decreased and an overall hit to the economy when taxes are increased. I agree that examining the empirical economic evidence is very important. But:

A. Uncontroversial empirical economic evidence is very hard to come by, since, in these matters, cause-and-effect is extremely hard to distinguish from mere before-and-after. Did the tax decreases two years ago really lead to the increased economic performance now, or is that instead just an accidnetal fact, as the increased performance is really due to those huge newly discovered oil reserves in northern France (or the revolutionary technological innovation in computer design, or the recent very good weather seasons [fewer hurricanes, and just the right amount of everything else for crops to grow extremely well], or ..., or ...)?

B. Even when there is uncontroversial empirical economic evidence, one can still question how economists make the value judgments necessary for proclamations of the "OVERALL better/worse sort". Is it just a wash (i.e., neutral), or rather good (or even bad), when some economic cause has the economic effect of causing 100,000 rich people each to lose $10,000 a year income, but 100,000 poor people each to gain $10,000 a year in income?

C. To the extent that I have any sense of what the empirical economic evidence has to say on the matter, it points only in the direction of CHANGES IN TAXES having effects: a decrease in taxes will give the economy a temporary overall boost, an increase in taxes (at least given how those tax revenues have traditionally been spent) will give the economy a temporary overall hit, but a CONSTANT HIGH LEVEL OF TAXATION is NO BETTER OR WORSE than a CONSTANT LOW LEVEL OF TAXATION. (My sense about this comes from my macro-econ class at College of Idaho as an undergrad. I'm quite confident that, minimally, my professor was being very honest [i.e., had no agenda in reporting such putative empirical economic evidence] since he was very explicitly and unabashedly an extreme financial conservative. In fact, for this reason, he seemed a bit unnerved when telling us about this.) So, as president, while I would increase taxes (at least for some individuals), to the extent that empirical economic evidence suggests any negative consequences at all, they will be temporary.

7. Perhaps the most common argument is: government simply spends tax money stupidly. I won't deny there is a large degree of truth in this. For example, if the government spends $1 billion funding a pointless bridge, then (although there WILL BE stimulation to the "bridge-building" part of the economy), to that extent, it will do very poorly by way of efficiency in satisfying needs/wants. But there can be no general, exceptionless argument of this sort. It depends on the composition of the government, i.e., who is in the government doing the spending. And I say to you: as president, I will spend tax revenues in ways that are VERY EFFICIENT when it comes to satisfying needs/wants, in ways that OVERALL stimulate the economy more/better, and in ways that OVERALL incentivize (rather than disincentivize).

By the way, I fully recognize that none of this is an argument FOR higher taxes. Instead, it's a way of saying that all of the most common standard, general arguments against higher taxes seem extremely problematic.

My argument for higher taxes will come in other places, where I talk about very worthwhile projects, etc. which can only get off the ground and succeed with funding that ultimately comes from tax revenues.

6 comments:

Nathan Hall said...

This is certainly the most thoughtful discussion of tax policy I've seen yet from a candidate. I have problems with point one, because there is what economists call a "dead weight loss" involved in taxation, so that it affects both which sectors of the economy are stimulated AND how much total productivity there is. I'll try to say more on that later. For now, I'd like to briefly express doubts about points 4 and 5.

Re (4), I agree that the rate of taxation does not impact the burden of transaction costs. However, the sort of system you seem to be advocating, where taxes are used to modulate the incentives of a free market toward better ends, is likely to feature relatively high transaction costs. The IRS may be getting close to the same number of returns every year, but it's much easier to process a flat tax than one with myriad provisions for utilitarian social engineering. In American politics, advocates of simple taxes are usually advocates of low taxes, so the distinct issues of how much money taxes should raise and how complex the tax code should be often get conflated. At least, so it seems to me.

In (5) you contend in part that disincentives for hard work could be mitigated by "'reverse taxation' for some individuals, i.e., adding some (relatively) fixed percentage to their income, for purposes of incentivizing." I do not think this would work, at least if you mean to incentivize hard work in one's job. Subsidies affect the market price of a product (in this case labor). The additional money offered as an incentive would just be compensated for in the market, as employers would offer lower salaries to prospective employees who would also be getting checks from the government if they accept the job. As a prospective employee, getting $80,000 a year from my employer is no different (in theory) than getting $60,000 from my employer and $20,000 in incentive pay from the government. Thus, the government cannot artificially incentivize jobs above their market price in this way, unless it directly pays the whole salary.

Nathan Hall said...

I think I was wrong about (5) a moment ago. A subsidy should raise the market value of the labor in question, but by less than the amount of the subsidy, the exact amount less depending on the elasticity of the market for that labor. Sorry, economics was a long time ago!

Brian Kierland said...

Thanks for the comments.

On (1), I'd like to hear more about "dead weight loss". So if and when you have more time....

On (4), excellent point, there are all sorts of ways that different tax systems can involve different levels of transaction and administration costs that have nothing to do with sheer number of tax returns. However, the conclusion of my argument was simply that higher taxation levels do not intrinsically involve higher levels of transaction and administration costs.

Of course, as you rightly note, there are many different ways that tax systems can differ besides something like overall/average rate. But I don't think anything I said, either in the original blog entry we are commenting on, or in other entries, commits me to anything other than a flat tax. (NOTE: saying this is importantly different from actually endorsing a flat tax.) We must distinguish the government's tax system from the government's spending system (expenditures, budget, however you want to put it). There are two steps: 1. the government gets money through taxes; 2. the government spends this money. The things I've proposed (or at least suggested as possibilities), to which you give the label "utilitarian social engineering", concern step 2, not step 1, and so they don't concern the tax system. Thus I still stick to my original claim.

What you say, though, does, of course, suggest a potential worry for the overall spending system that I will (far from completely) be describing across my many blog entries. This worry is that such a complicated spending system involves unavoidable transaction and administration costs. I concede to this worry in the sense that, yes, such costs are unavoidable. The question is whether they are worth it, and that would presumably have to be decided on a case-by-case basis. (I assume, though if you question please share, that there are no "emergent" transaction or at administration costs that only "emerge" at the level of whole spending systems.) Making really firm judgments about each case would require a higher level of detail than I am (or any candidate is) attempting to provide. But I think it is quite plausible that a least some kinds of government expenditures are well worth the transaction and administration costs. (If you disagree, let me know, and I will try to defend.) Thus, I'm not worried about any sort of general argument against complicated spending systems of the sort that I propose.

On (5), glad to hear that you have corrected yourself.

Anonymous said...

Professor Kierland,

Congratulations on your campaign. I have enjoyed reading many of your blog entries. Most seem very pragmatic. I can’t claim to fully understand them all, but I think I follow a good number of them.

Among your writings that caught my eye is point #5 under “taxes.” Here you state, “…there can be no general argument that taxation necessarily disincentivizes OVERALL.” This seems right (though I lack the intellect and skill needed to prove it.)

Still I think an argument that taxation does disincentivize overall might be made general enough for all practical cases given a utilitarian-minded government as I think you propose.

First, as a definitional point, I assume we agree that if we increase disincentives overall, then we would expect aggregate production and income to decrease, all else being equal. Similarly, if we increase incentives overall, then we would expect aggregate production and income to increase, as well. (If I have somehow muddled your meaning of incentives by linking changes in aggregate incentives to changes in aggregate production and income, please stop me here.)

Now, if I understand your argument correctly, you effectively claim it is not the case that a targeted tax will necessarily lead to lower incentives (and hence lower production and income) overall. This is because the tax revenue can be used as incentives to others who will boost their production to gain more income, possibly even to the point of matching or even besting the levels of aggregate production and income before the targeted tax was imposed.

So far do I follow you correctly? If so, there is no argument from me yet.

Now, in your response to Nathan, you clearly separate the questions of how government imposes taxes from how it spends taxes. But the former question needs consideration, I think. While I am not clear on what, if any, tax system you have committed to, I think your utilitarian platform may imply commitment from you to at least some degree.

In particular, I suggest a government attempting to maximize benefit among the greatest number of individuals will seek to tax higher income earners first. I think this would be so if we accept that the nth dollar earned offers less marginal benefit to a given income earner than the first dollar earned by another income earner.

What’s more, higher income earners tend to have a greater “negative taxation elasticity of production” than lower income individuals. That is, higher income earners, on average, will choose to reduce their production a greater percentage than lower income earners in response to some percentage increase in taxes. Why? Simply put, because higher income earners are the ones who can afford to do so.

Now, after collecting a tax, government must spend the tax on production activities that are more efficient (in terms of production per dollar spent) than the production activities that would have occurred had the persons taxed been allowed to consume or invest the amount taxed, else production and income will be less in the aggregate compared to what they would have been without the tax. This is so because the tax spent must result in production that offsets 1.) the next-period production lost from the taxed income no longer available to the income earner for consumption or investment, and 2.) the additional current-period production lost due to the disincentive effect of the tax (which is maximized by a utilitarian-minded government that we can assume will tax higher income earners first due to the diminishing marginal benefit of income.)

However, it seems implausible that government will be able to find more efficient production activities to fully offset 1.) and 2.) because the higher income earners tend to be the same individuals who have already identified the most efficient production activities. That is, they earn higher incomes precisely because they are more efficient producers. But our targeted tax collection is likely to encourage less production within these more efficient production activities, not more.

So, given the assumption that a government seeking to maximize benefit among the greatest number of individuals will tax higher income earners first, I think we can establish that taxation will result in greater disincentive overall, and thus lower production and income overall.

I understand a utilitarian-minded government may not be as concerned with aggregate production and income as it is with the distribution of income among all individuals. Such a government may be willing -- even eager -- to accept lower total income provided it leads to higher median income. (And I accept there may be numerous tax-and-spend strategies able to accomplish this.) But this seems to be a different argument than saying that taxation does not necessarily disincentivize overall -- again assuming that taxation is purposefully targeted at those who stand to lose the least marginal benefit, i.e. higher income earners.

Thanks for your consider,

David Fairchild
Olathe, Kansas

Nathan Hall said...

You seem to believe that the nth dollar earned by one income earner has more marginal benefit than the mth dollar earned by another income earner, given m < n. Do you think this is generally so? Especially given your views on the singularity, isn't it possible that an extra dollar in the pocket of a well-to-do person could give them immortality, while the mth dollar of someone who makes ends meat but little more is relatively inconsequential?

Dave said...

Nathan,

Thanks for your question. I am not sure if it is directed at me or at Professor Kierland. While I talked about the notion of declining marginal benefit of the dollar in my earlier reply, he discussed the singularity and immortality in detail in another of his blogs.

I do think that generally there is declining marginal benefit of income, just as there is declining marginal value of virtually any economic good I can think of. (By economic good, I mean something that has both utility and scarcity.)

When it comes to "buying" more life, I think of healthcare. And today, at least, I think we can easily show that each extra healthcare dollar spent by an individual will buy less and less marginal healthcare benefit. For instance, if we expand the definition of healthcare a bit to include very basic health-related purchases, we might take into consideration sanitary water systems that can save many lives for relatively little cost per person. Contrast this with very expensive heart disease or cancer treatments that might only extend the life of a very sick person a few months on average.

The singularity idea, if I understand it correctly, I think would, in fact, change the study of economics forever because it would suggest there is no longer declining marginal benefit of the dollar, at least when it comes to the purchase of healthcare.

Healthcare would become the first and only god I know of that would defy the law of DMB in a sustained manner. (There are previous notable short-run defying examples [such as the often-cited Irish potato famine], but I can't think of any long-run examples.)

Thanks.