Revenue Committee Uses Flawed Economics

In its Thursday meeting, the Revenue Committee approved a bill to introduce a sales tax on services while lowering the overall state sales-tax rate (scroll to 1:20:00 in the video below).

This bill is part of a tax policy strategy laid out by the Revenue Committee, based on two terms: "revenue neutrality" and "constant revenue". The first term was used frequently in the committee as it discussed the bill, known formally as 19LSO-0232, to describe the tax reform that the Committee envisions.

The term "constant revenue" has been tossed around for years by legislators and tax-hike pundits. It is used to explain the need for tax reform: government should not have to worry about fluctuations in tax revenue due to fluctuations in the business cycle.

Unfortunately for the members of the Revenue Committee, the problem is that neither term is meaningful from a policy viewpoint. Outside fancy economic theory there is no such thing as "revenue neutrality". As for "constant revenue", short of central economic planning, it simply does not exist.

Revenue Neutrality

The concept of revenue neutrality refers to the idea that you shift taxation from one economic activity to another. For example, if a sales tax only applies to goods, and not services, a revenue-neutral tax reform lowers the tax on goods while extending taxation to services. The idea is that government will collect as much in tax revenue after the reform as before.

I have studied tax reforms off and on for a good 25 years now, and I have yet to find a single one that meets the goal of revenue neutrality. This is not surprising, and there are two reasons for that.

First, the very idea of neutrality is flawed. Those who talk about revenue neutrality never really attempt to define when, where and how their tax reform is going to be revenue neutral. The reason is, quite possibly, that they do not bother to distinguish between static and dynamic neutrality. Static neutrality means that the new tax system, after the reform, brings in the same amount of tax revenue as before, at every level of economic activity.

Dynamic neutrality, by contrast, means that post-reform taxes bring in the same amount of tax revenue over the course of one business cycle, as pre-reform taxes did.

When I listen to those who suggest a revenue-neutral tax reform here in Wyoming, I get the impression that they think of neutrality in static terms. For that to happen, they have to assume that their reform will not change economic behavior among taxpayers, an idea that drifts farther away from economic reality the more complex the reform gets.

A tax reform that is very simple and very small will have no visible impact on economic activity. For example, a one-point cut in the sales-tax rate on whole milk and an increase of one point in the sales-tax rate on 2-percent milk will pass undetected by consumers. By contrast, a tax reform that lowers the taxation on goods by 0.5 percentage points - as the Revenue Committee discussed - but creates an entirely new sales tax on services, will be big enough to alter economic behavior. Therefore, it is neither theoretically nor empirically meaningful to talk about static revenue neutrality.

Econometricians will disagree and claim to be able to pinpoint exactly what shifts there will be in consumer behavior in response to the tax reform. I invite them to try. Again, I know of no single tax reform that has ever met any neutrality test, despite confident pre-reform claims from econometricians. But by all means - give it a try...

Neutrality advocates also try to approach static neutrality from the viewpoint of the consumer's bottom line. They suggest that people will save enough money on the sales-tax cut on goods to make up for the increased cost for buying services. Therefore, they say, people will have enough money left in their pockets to continue to buy services the way they do now, only with the new sales tax in place.

However, this argument changes the very premises of their reasoning. Suddenly, neutrality is no longer a matter of revenue, but of purchasing power. By suggesting that taxpayers will not see a difference in their checkbooks, tax reform proponents hope to improve their sales pitch of the reform to skeptical voters. What they forget, though, is that this is no longer a tax-neutrality argument. It is an argument on income neutrality. But even as such, it still does not validate any argument for a tax reform, and the reason is simple: if you save half-a-percent in sales tax on shoes, clothes, gasoline and books, there is nothing that says that your regular purchases of services will be such that it exactly, or even approximately, makes up for a 3.5-percent state sales tax on services.

Furthermore, the income-neutrality argument is based on the premise that the state sales tax is the only sales tax that will be added to services. However, once the state starts collecting sales taxes on haircuts, taxi cab rides, plumbing, real estate sales, auto repair, truck transportation, legal services (oh, wait, I think the Committee explicitly decided to exempt legal services; I wonder why?) there will be local sales taxes on those services as well.

This means that
  • you will pay 0.5 percent less on goods, and
  • you will pay 5-6 percent more on services. 
Long story short: the sales-tax reform suggested in 19LSO-0232 is not income neutral.

It is also not revenue neutral from the static viewpoint. But what about dynamic neutrality?

This version of revenue neutrality rests on the idea that if a tax system brought in $100 in revenue over a business cycle before a tax reform, it will bring in $100 in revenue over a business cycle after the reform as well.

Again, we are dealing with a nice theory, and in fairness to the neutrality advocates, this one actually stands a little bit of a chance in the real world. The problem here is that it hinges on an implicit axiom that is never spoken of: that the business cycle looks the same before and after the tax reform.

Most of the economic theory that is used by modern-day economists became formalized in the mid-20th century. Back then, business cycles were smooth and easily predictable, with about four years in total over one growth period and one recession. The peaks and troughs were also predictably smooth. Based on what they saw, the economists of that era wrote macroeconomic textbooks - and developed the theory that underpins modern-day econometrics - that explained business cycles in precisely this smooth, predictable manner.

There is economic theory that does not rely on this standardized explanation of the economy. However, such theories, which range from real business cycles to Post-Keynesian economics to Austrian theory, are not what most econometricians use. There is a sliver of formalized models that rests on real business cycle theory, but I have yet to see it successfully applied in the field of tax reform. It may exist, and if it does, I will be happy to take a look at it.

However, the application of a non-mainstream theory would not save the dynamic neutrality idea. On the contrary, it would blow a hole in the very argument: the common denominator for all non-mainstream macroeconomic theories is that they part with the smooth business cycles of mainstream macroeconomics. Therefore, they do not even allow the assumptions needed for dynamic revenue neutrality.

In short: the only way anyone can make a logical argument that a tax reform will be revenue neutral over a business cycle, is if he assumes that the business cycle remains unchanged from what it was like before the reform. But since the late 1960s, the business cycle in the U.S. economy has changed profoundly:
  • In the 1970s we experienced stagflation, with inflation and economic stagnation at the same time;
  • In the 1980s and 1990s we had two long growth period with strong expansion in the 3.5-4.5 percent range of annual GDP growth;
  • In the 2000s we had a ho-hum growth period that barely generated three-percent growth;
  • The recovery from the Great Recession was a macroeconomic embarrassment, without a single year of three percent growth.
The business cycle here in Wyoming has been at least as erratic as the national business cycle. Therefore, without a very careful, scholarly astute thesis on the subject, it would be foolish to the point of macroeconomic malpractice to suggest that the next business cycle is going to mimic the last one.

Plain and simple: it would be highly irresponsible to base a tax reform here in Wyoming on either concept of revenue neutrality.

Constant Revenue

This concept has two dimensions, a moral and an economic. The moral dimension is the idea that government has some kind of entitlement to a constant stream of revenue that does not fluctuate with the business cycle. In other words, it takes the concept of revenue neutrality to an even higher level.

It is, of course, both cynical and egoistic to argue that government should stand above the ups and downs of the business cycle. We the taxpayers have to live with the ebbs and floods of growth and recessions; what makes government more important than you and me?

Things do not get better when we move to the economic dimension of the "constant revenue" concept. To make it work in practice, the legislature would have to create a regressive tax system, probably the economically most destructive tax system a government can come up with. 

To understand what a regressive system is, let us first start with a regular system, such as the one we have today (and practically every other functional economy has). Generally, taxes are a proportional share of the value of economic activity, such as a sales tax that grabs a certain, fixed percentage of what we spend on the taxable item. A flat income tax, or even one that is only mildly progressive, is also proportionate in claiming a stable percentage of our income. Even a property tax is proportionate, provided property values do not change much more than incomes do (a good generalization but less so than the one about income taxes). 

Taxes on businesses also fluctuate, by and large, in parity with the fluctuations of economic activity. The benefit of this is that businesses have a reasonably good chance to predict the cost of taxes over time. 

To government, however, this is not a benefit. When tax revenue remains in parity with economic activity, it also declines with a decline in said activity. Government does not want to cut its spending - there are school districts where superintendents lament the loss of two administrative positions while in that same county a thousand private-sector jobs are lost for every lost school administrator job - which has led some elected officials to call for "constant revenue" tax reform. 

Unfortunately, there is only one way for government to construct a constant-revenue tax system, and that is to create regressive taxes. Here is how it works, in aggregate:

1. In year 2020, the Wyoming GDP is $100. Taxes bring in $10 in revenue for government.
2. In 2021, there is a recession. GDP falls to $95. Under the present system - simplified of course - tax revenue declines to $9.50.
3. We now roll back time to 2020 again and decide to introduce a tax system that will prevent the $.50 tax revenue loss.
4. Since GDP declined, there is less economic activity to tax. Therefore, the only way to secure $100 in tax revenue under a GDP that is only $95, is to raise taxes. 
5. To make a tax system operate as a constant-revenue system, the legislature would have to build in automatic tax hikes, so that when a business loses sales, the tax rate on its activity - sales, use, excise or even gross receipts - would have to go up.

This last point is crucial. In order to guarantee constant revenue in the aggregate, the legislature will have to guarantee constant revenue in the disaggregate. Suppose a book store sells books for $1m in 2020 and delivers $60,000 in sales tax revenue. In 2021 sales drop to $900,000 but government, in its desire to keep tax revenue constant, still wants $60,000. Therefore, the sales tax will automatically have to rise from six percent to 6.67 percent. 

Property taxes effectively work this way: the value of the book store's property is not re-assessed because its sales dropped. However, the tax base is not its sales either, but a variable - the property value - that is independent of sales, at least over the short term. The problem with making the sales tax a constant-revenue tax is that it is not possible to predict sales to a point where you can raise the sales tax beforehand. How would you know what the rate should be?

To get around this problem, the legislature would have to reconfigure our tax system. They have two options:

To move entirely to a tax system where the tax base does not fluctuate in value with the business cycle, or 
To rely entirely on taxes that are independent of the business cycle.

The former alternative runs into the problem exemplified by the sales tax above. Income taxes are plagued with the same problem; if a person loses his job, is he then supposed to continue to pay income taxes as if he still had his job, just to guarantee government a constant stream of revenue?

The latter alternative means, strictly speaking, a lump-sum tax owed by everyone regardless of employment status, income, spending, saving, wealth or other taxable status. Margaret Thatcher pushed a tax like this through the British parliament. It was called a "poll tax" and did not last long, for easily understandable reasons. Her motive was not constant revenue, but the reliance on flawed economic theory. However, I doubt a poll-tax style tax would even make it through the Wyoming legislature.

In lieu of a theoretically perfect constant-revenue tax, the legislature would have to opt for a radical expansion of its reliance on property taxes. Property values are much less volatile than income, consumption and business sales. However, even if our elected officials decided to shift taxation away from economic activity and onto property, it is highly unlikely that they could guarantee themselves and government a constant-revenue solution anyway. They would, namely, have to make the radical transition from today's system onto a property-tax reliant system under the auspices of "revenue neutrality". As mentioned earlier, the more radical a tax reform is, the less likely it is to be revenue neutral. 

Even more important is the problem that a regressive tax system causes for economic stability. If property taxes were substantially higher, they would claim a substantially larger portion of people's incomes than they do today. The termination of sales, use and excise taxes in proportion is unlikely to have a neutralizing effect, especially since the shift in taxation is not individual-specific. Many property-owning households are likely to experience a loss in net-tax purchasing power in the first place, a problem that will exacerbate the regressive nature of the tax system.

In other words, when the economy goes into a recession and the new constant-revenue tax system is supposed to maintain tax revenue, the high property taxes will erode declining household income more quickly than they do today. More people will default on their mortgages and their property taxes earlier in the business cycle. More people will lose more money more quickly and therefore have less to spend. Consumer spending declines faster, earlier and probably overall more steeply. The recession accelerates, foreclosures rise - and property tax revenue starts tumbling. 

Even if a constant-revenue tax system is not designed specifically to rely on property taxes, its design has to follow the sketch laid out here with the property-tax example. Tax rates have to rise during a recession, thus aggravating the recession itself. The business cycle becomes more volatile.

Some proponents of a constant-revenue system suggest that we can find a tax system that relies on industries that do not move in tandem through the business cycle. Again, this is a nice theoretical idea but in practice, no such tax system exists. There are many reasons why, the main one being the aforementioned point about the nature of the business cycle: we cannot predict with the necessary level of confidence what the next recession is going to look like.

One could argue that we can reform taxes from one business cycle to the next, in order to try to anticipate shifting patterns in the economy. However, this would inject yet another element of uncertainty into the system (I happened to write my doctoral thesis about precisely this form of institutional stability...) Since a tax system must be institutionally stable to even meet basic criteria of decency vs. taxpayers, the legislature cannot shift taxation around from one recession to the next. Therefore, when we decide to make changes to the tax system, we need to make them rare, predictable, transparent and economically as harmless as possible. 

This rules out both revenue neutrality and constant revenue as criteria for any tax reform. 

Instead of obsessing with keeping government funded at any cost to taxpayers, our legislators might want to start looking at the other side of the equation. 

What side is that? Oh, that's right.


No comments:

Post a Comment

Weekly Economic Review

Our lawmakers here in Cheyenne are planning to raise taxes on the tourism industry - by a lot. In the meantime, the first issue of my new w...