Recent Posts

Wednesday, April 16, 2014

The Debate Over The Purpose Of Taxation

Garth Brazelton's article "The Purpose of Taxation" discusses the purpose of taxation, and this has generated a fair number of comments on the Mike Norman MMT website. His article discusses government finance from the point of view of Functional Finance. In this article, I want to give a high level discussion of the controversy around the Functional Finance viewpoint on the role of government taxes.

Garth Brazelton's key argument is as follows:
So what purpose does taxation at the federal level serve?  Well, largely it's to modulate private spending demand.   (I would argue a secondary but significant purpose is income redistribution to regulate economic inequality)  When taxes rise, resources are taken from me, Indiana, and other entities that have limited funding sources because we don't control our own currency.   Because of that, our spending demand shrinks.  Contrarily, when taxes fall, resources are given to me, you, and Indiana, and our demand to spend some or all of that money rises.  The fact that the government has changed the taxation level at the federal level need not have any bearing whatsoever on the federal decision to spend - it effects the private decision to spend greatly however, and in that way, regulates spending demand and inflation.   This is a fundamental difference between the US government compared to individuals and States (and even the European Union countries like Greece that also do not have control of their own individual currency).   
This is a restatement of Functional Finance, which was developed by the economist Abba Lerner (for a longer discussion, see my primer).

However, in the comments Ramanan countered:
Imagine a economy with the government expenditure about $15 and average tax rate around 15% and the GDP around $100. And another with $20, 20% and $100. And both at near full output. For the first to have government expenditure at $20, it may have to wait for a while for productive capacity to go higher while the government increases expenditure from $15 to $20 over time. The other route is increase average tax rates, so that the government can spend more. So in that sense taxes fund expenditure. There is no need for making counter intuitive claims that taxes do not fund expenditure and so on. [emphasis added -BR]
In other words, the Modern Monetary Theory version of Functional Finance is just "word play".

What Are The Operational Differences?

I do not think the observation that taxes do not "fund" expenditures is just word play. However, I do not think verbal argumentation will lead anywhere, rather we have to look at the operational differences between the points of view.

I now return to the article "Functional Finance vs the Long Run Government Budget Constraint" by Nick Rowe (which he wrote almost exactly three years ago, when this dispute previously rolled over the blogosphere).

Functional Finance says you only use taxes if you want to reduce Aggregate Demand to prevent inflation. The Long Run Government Budget Constraint says you use taxes to pay for past, present or future government spending. They sound very different. They aren't.
There's a general principle in economics: first you eat the free lunches; then you  look at the hard trade-offs. Functional Finance says "first eat the free lunches". The Long Run Government Budget Constraint says "then look at the hard trade-offs".
That's the underlying kernel of truth in Abba Lerner's Functional Finance (pdf) [see also his book The Economics of Control (pdf)]. And I don't know of any mainstream macroeconomist who would disagree. [emphasis added - BR] You use taxes only so you don't have to print money. When you get to the point that Aggregate Demand is high enough, so you start to worry about inflation, you use taxes to finance past, present, or future government expenditure precisely because you don't want to print more money and make inflation higher.
The important dispute is about the long run government budget constraint, not the legalistic details of monetary operations. If the long run budget constraint holds, then taxes have to be used to pay for spending - eventually. And the mainstream view is that the budget constraint is binding.

The economists who drove the development of Modern Monetary Theory reject the long run budget constraint, as they argue that the assumptions are too restrictive. (For example, Bill Mitchell on The Billy Blog has written extensively about this.) I am in agreement; the long run budget constraint to my eyes looks to be either trivial or wrong. It does not act as a binding constraint within models that properly model fiscal policy within a welfare state.

And once the governmental budget constraint is rejected, the concept of taxes paying for spending is nebulous. There is a large gap between spending and taxes, and the gap grows in line with the economy.

Importantly, the effect on demand depends upon the form of the tax. Although many free market economists will probably disagree, I argue that taxing poor people has a much greater impact on demand than taxing the rich or corporations. If the weight of the aggregate tax burden was shifted back towards corporations and the wealthy*, the government would need to raise more money in order to have the same effect on demand. (And Japan is about to find this out the hard way with its sales tax hike - again.)

Therefore, the government cannot judge its fiscal stance solely by the amount of money it raises by taxes, it also has to consider the mix of taxes. Such a statement world be hard to reconcile with a simplistic view where the role of taxes is solely for financing expenditures. As such, I do not view the argument as being just a question of semantics.


* Although income taxes are mainly paid by the rich, payroll taxes (such as Social Security) are an extremely important de facto tax that are mainly paid by the broad working population. Deductions for government pensions are only disguised taxes, as I discuss here.

See Also:
(c) Brian Romanchuk 2014


  1. I couldn't agree more with you here Brian:

    "f the weight of the aggregate tax burden was shifted back towards corporations and the wealthy*, the government would need to raise more money in order to have the same effect on demand."

    If tax cuts for the wealthy do not stimulate the broader economy to the upside, then why would we expect tax increases on the wealthy to diminish broader economic activity aka have a strong deflationary bias?

    If taxes serve only 3 purposes:

    1) give fiat currency a baseline value and level of acceptance
    2) regulate aggregate demand
    3) punish in order to minimize undesirable social actions (pollution, smoking, trans fats, excessive wealth)

    Which one of the 3 purposes does increasing taxes on people making $250K per year serve (the original fiscal cliff tax increase baseline for Dems in 2012)

    what about 90% tax rates on incomes over $20million per year? Which purpose does that serve?

    Amazing how much the optimal policy choices change depending on the framework one is using.

    1. Attempting to reduce inequality through the tax system has been supported by a fairly wide range of political thinkers. I am in that camp; my political views are nonstandard, but I fit somewhere in the pre-1980s consensus. That said, I do think that our current mixed economic system does have innate tendencies to widen inequality, and those tendencies are very difficult to counter. This means that I do not think it is a good idea to attempt to push the tax system too hard in an attempt to flatten inequality.

      If we tax incomes above some threshold of $X with a marginal rate of 90%, what will happen is that tax return data will show very few people earning above $X. But does that mean that we suddenly have flattened the income distribution, or has the income shifted into non-taxable forms?

      What we saw in practice with the very high marginal rates after the war is that the tax system ended up full of loopholes. It is difficult to sustain a marginal tax rate over 50%.

    2. I agree with everything you say here, I was just illustrating that I agree with you that the deflationary impact of any given tax increase has to do both with its relative size and with who is paying the tax. I was just trying to frame it in another way.

  2. Nice post, Brian.

    Three quick points. First, Blinder and Solow showed long ago in "Analytical Foundations of Fiscal Policy" (1973) that the balanced budget rule (or any other similar fiscal rule, such as Maastricht or whatever) are deficient. They said back then that "it's the economy that needs to balanced" (they were thinking of several issues, including the public's saving desires, etc). Interestingly, the same article also shows that federal spending via bond financing has a greater long run multiplier than the same amount of spending done via money financing. And yes, they also mention the corollary: that open market purchases are contractionary in the long-run...MMT should take note.

    Second, no matter how you look at it, the statement "taxes fund federal spending" holds. Taxes fund federal spending in real terms (in terms of its utilization of productive capacity in the economy), just not in nominal terms. I would advise against saying that "taxes don`t fund anything". I actually agree with Ramanan on this. There is no need to frame it differently. If you want to highlight the issue of solvency, all you need to do is say that a government that issues its own currency cannot default. Period. (Blinder says that too)

    Third, (Old Keynesian) Robert Eisner in "The Misunderstood Economy" argued that the most important function of taxes is to regulate aggregate demand. However, he understood that under the current US institutional and governance regime, regulating demand via that mechanism is highly ineffective. He used to give the example of the 1967 (or 68) tax increase, which came into effect too late to kill inflation due to Congressional delay.

    1. With respect to the "taxes don't fund anything" point, I had noted previously that I am not comfortable with the extreme versions of that statement. But it is a fairly vague area. For example, a very small government could reasonably fund itself entirely by seigneurage, if it strictly enforced reserve ratios. Thus, theoretically, a government could fund itself completely in the absence of taxes. However, that would not work for a welfare state.

    2. Yes, that makes sense. BTW, my comment wasn't directed at you. I had just finished reading the post by Garth B and was making a general comment on the meme as sometimes used on blogs.

  3. Taxes are a method of assigning cost for government behavior.

    Funding cost by government borrowing or printing is a way of transferring property with indefinite but traceable wealth consequences. Both borrowing and printing increase the stock of financial property but only to the benefit of those in the growth chain. Those not in the growth chain see a dilution of their financial property.

    If we can accept that these are the ONLY two methods available for funding government, we can debate which method will be used, or better, what combinations of the two methods will be used. With either method, it takes productive people working long hours to produce real physical results.

  4. This comment has been removed by a blog administrator.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.