Periodically, the prospect of a new ‘wealth tax’ re-emerges on the platform of public debate. Currently, the Davis Tax Commission has invited submissions from the public on the ‘desirability’ and ‘feasibility’ of a potential wealth tax. Let me answer that real quick:

Desirability? The only people that would want it are the people that it doesn’t affect negatively. So, not taxpayers. Tax practitioners, however, would be in for boom times (see ‘feasibility’ below).

Feasibility? You must be joking. If you’ve ever witnessed the laborious process of trying to ascertain the value of assets and liabilities in a deceased estate, you’d know that trying to do that on an annual basis is an exercise in futility. If wealth includes houses, valuables, personal use assets, financial assets, entrepreneurial ventures, share options, money in bespoke investment schemes and the value of cash-in-able pension fund savings, less all the mortgages, credit cards, private loans and contingent liabilities associated as-yet-unascertainable-losses, then SARS will be looking at a mammoth administrative task. To be clear, a wealth tax is not as inherently simple as ‘money paid by company into your bank account’ or ‘VAT charged on all goods sold through your store’. SARS will find itself mired in a morass of speculative valuation. And all for quite a low return on their time investment. Because any tax practitioner worth his fee is going to become very conservative in his hunt for asset valuations, and just as conservative with the valuation of liabilities*.
*To be clear, conservative asset values are low, while conservative liability estimates will be high.

And this ‘feasibility’ problem is part of the reason why wealth taxes have fallen out of vogue. Half of the OECD countries had some form of a wealth tax in 1990. Today, only France and Norway have a wealth tax, and Switzerland in some of the cantons*.
*Thanks moneyweb for that statistic. Full disclaimer: much like Huff Post ZA, I haven’t fact-checked it. Sorry.

But let’s put the desirability/feasibility question to the side, and address something more important.

Because the Davis Tax Commission is pretending like the income tax is not a wealth tax. Oh, but it is.

In fact, I’d say that, if you’re into this kind of thing, it’s the most important kind of wealth tax that there is.

All Taxes are a ‘Wealth Tax’

Let’s start with two definitions:

What is wealth?

The standard answer: wealth is your assets, less your liabilities, which gives you your net assets.

And what are assets?

The (also) standard answer: assets are things that either earn you income (eg. shares will earn future dividends), or save you money*.
*which is just another form of income – e.g. owning a car means that you don’t have call an Uber every time you travel, which saves you time and money, which increases your disposable income.

So bearing those defintions in mind, I’ll give you two scenarios:

  • Scenario A: I own R300,000 worth of government bonds, earning 8% interest per year. That is: R24,000 in interest income will be deposited into my bank account every year, or R2,000 per month.
  • Scenario B: I receive a monthly government grant of R2,000 per month.

What is the difference between them?

In both situations:

  • the person paying the money out is identical;
  • the amount of income received is identical; and
  • the likelihood of “default” is about the same, because suspending government grants is about as politically poisonous as defaulting on government debt.

The only real difference is: if I owned the bonds, then I could sell them, take the R300,000 of capital, and do something else with it (either invest it or consume it).

When is an income tax the equivalent of a wealth tax?

Well, let me twist this story around a bit, and ask what return the bonds should earn before a normal person would never ever sell them? As a guess, I’d bet that it’s somewhere around 20% – especially if that return is going to be indefinite.

So scratch Scenario A – because it’s not really comparable to the government grant. And I’ll introduce a third scenario instead:

  • Scenario B: I receive a monthly government grant of R2,000 per month.
  • Scenario C: I own R120,000 worth of government bonds, earning 20% interest per year (R24,000 a year, or R2,000 per month). The only catch is that I’m never allowed to sell them, which is why they earn such a high return.

At this point, those scenarios are effectively identical.

Or put differently: a monthly government grant of R2,000 is a R120,000 asset*, redistributed indefinitely.  Perhaps we can quibble about just how high the return must be in order for you to never sell the bond. But the real point is that a regular government grant is a kind of asset, because an asset is really just a bundle of expected future incomes.
*In technical terms, that government grant represents a R120,000 claim on the total assets held collectively in the fiscus.

And essentially, the income tax turns taxpayers into asset managers on behalf of net welfare beneficiaries.

But then it begs the question “Why do you need a special wealth tax?” Because every tax already does that.

Taking that back to South Africa’s Numbers

I want to go back to a table mentioned in an older post: How Taxes Reduce Inequality in South Africa. It splits the population of South Africa into 10 parts, and ranks them from poorest to richest. The table then takes each decile’s average market income, deducts direct taxes, and finally adds back cash transfers (welfare grants) to work out average disposable ‘post-fiscal’ income. Think of it as: the amount of money that actually ends up in people’s bank accounts.

Screen Shot 2015-10-27 at 7.22.23 AM

If you break down the difference between ‘market income’ and ‘post-fiscal income’, you get something like this:

Reallocation of wealth through an effective wealth tax in South Africa
“Net Transfers” are the effective benefits (positive) or taxes (negative). “Cash Transfers” are tax money that are directly transferred into poorer people’s bank accounts. “Contributions to Free Basic Services” are the tax money used to pay for government-provided services (eg. defense, subsidised healthcare, public education, etc).

But that’s only per person. So I’m going to increase that to the population size per decile, and rework those cash transfers into their representative asset terms (based on that 20% return):

Screen Shot 2015-11-04 at 12.19.56 PM

Basically, in terms of cash transfers alone: taxpayers are effectively expected to put aside a R207 billion portfolio of their own assets, and directly accrue that income for the net beneficiaries of social grants.

But remember that this is valuing the asset from the perspective of the cash grant recipients. If you’re the taxpayer, you can’t expect to earn a 20% return on those ‘set-aside’ assets. And in the real world, where real returns might be closer to 3% (in SA), we’re talking about an effective carve-out of around R1.3 trillion worth of taxpayer assets.

For the record, the difference between those two figures is essentially value destruction – where the value of the cash transfer is less in the hands of the recipient than it is in the hands of the giver. Or more colloquially:

Recipient: “This R2,000 per month is paltry!”

Taxpayer: “How dare you ask me to give up R800,000* of my own wealth in perpetuity!”
*that’s the investment that you’d have to make in order to hypothetically pay out R2,000 per month in perpetuity, earning a 3% real return.

Total Tax Collection as an outsourced Asset Management Program

The example above is only referring to direct cash transfers (ie. social grants). What would happen if we were to include the value of the “Free Basic Services” that form part of government spending?

To measure that, we could take the R482 billion of personal income tax that the government plans to collect in 2017 and value that as an indefinite income stream.

Using that 20% rate of return, we have an effective R2.4 trillion asset management program.

That is: taxpayers collectively are expected to put R2.4 trillion of their collective assets to work for someone else.

And to put that into perspective, the 2014 World Ultra Wealth Report estimated South Africa’s total wealth at USD 630 billion. That’s just over R8 trillion.

So how is personal income tax anything other than a long-term, indefinite wealth tax?

PS: I’m not saying that the current situation is wrong, and I’m not saying that I disagree with the current tax regime. I think that most of us would agree that unequal societies require redistribution. But I am saying that if you’re going to address wealth inequality, or income inequality for that matter, the you need to take into account the existing policy measures that correct for them.

PPS: I’m also aware of the fact that many of these figures are debatable. But the numbers aren’t all that important. The point is that there is a principle here. ‘Assets’ are just future income. If you’re going to tax future income, then you are already taxing the asset. And therefore, we shouldn’t really talk about ‘wealth taxes’ as somehow different to ‘income taxes’ in any principle-based way. The only real difference is a practical one: a new wealth tax would be a form of double taxation. Just like donations tax and estate duty are already.

Rolling Alpha posts about finance, economics, and sometimes stuff that is only quite loosely related. Follow me on Twitter @RollingAlpha, or like my page on Facebook at Or both.