I am expecting some  modern monetary theory  (MMT) directed criticism for having signed and posted the letter from a wide range of activists, practitioners and economists to the G7 on the need for a financial transaction tax to help fund vaccine programmes in developing countries. This blog post is an explanation for why I have done so.

There are a number of important framings for this argument. The first is that I think there are six reasons for tax:

1) To ratify the value of the currency: this means that by demanding payment of tax in the currency it has to be used for transactions in a jurisdiction;

2) To reclaim the money the government has spent into the economy in fulfilment of its democratic mandate;

3) To redistribute income and wealth;

4) To reprice goods and services;

5) To raise democratic representation - people who pay tax vote;

6) To reorganise the economy i.e. fiscal policy.

At best most people who support MMT only recognise 1,2 and 6 here. That is a shame. It puts many who support MMT in a very unfortunate policy arena. To deny that tax plays these other roles, and that they are important, is a simple mistake: there is no other way to describe it.

Second, we need to consider the role of tax in government funding. I have published on this a number of times. This article in the Real World Economic Review is readily accessible so I will use that as my reference point. As I argue there:

As Mitchell et al (2019, 333) suggest, within MMT the macroeconomic identity describing the monetary funding of government expenditure (G) can be summarised as follows, presuming T is the sum total of taxes raised in cash during a period, B is government borrowing and M is government created money, with ∆ representing the change in a total during a period:

G = ∆B+ ∆M + T

That, I suggest is an identity: it must hold true in a country that does not receive development aid. In that case, if D is spending on development aid in a country and R is non-development spending (so that D+R = G) then in a country spending on development aid:

D =∆B+ ∆M + T - R

D here is, of course, a spend.

The first noted identity is changed in countries in receipt of this aid. It becomes:

G = ∆B+ ∆M + T +D

Here D is income.

So the question is why might development spend need to be funded by tax?

One is simple refusal to increase borrowing or increasing money creation. Another is refusal to reduce other spend. Although glaringly obviously not true a government may argue that ∆B, ∆M and R are at their limits. This is not true in any developed country at present. It does not mean that the argument will not be used.

There is another reason. That is that the economy is at full employment. Additional spend does require that space be created within the economy to permit additional government spending on a productive activity, on this occasion vaccine manufacture, if inflation is to be avoided.  This is an MMT consistent argument as to why tax must be used. Again, this not true in any developed economy right now. But that does not matter in political terms: the argument is being made that there is an inflation risk at present and in that case this issue kicks into play.

So although MMT says there are other options for developed countries when funding development spending, politically they are not feasible.

The developing country that is in receipt of aid also needs to be considered. Remember that for MMT to work there are pre-conditions:

• The jurisdiction's own currency is mainly in use in the country;
• There is a functioning central bank;
• There is no borrowing in foreign currencies;
• There is a functioning (and not corrupt) tax system;
• The tax take is big enough to have an influence on economic policy: if the yield is too small the impact is too low.

These conditions are rarely met in developing countries.

So, turning to the formula for these countries:

G = ∆B+ ∆M + T +D

Borrowing imposes burdens, often expressed in US dollars. That is undesirable when no additional capacity to pay is being created.

There is no effective mechanism to recover additional money creation from the economy. This, then, would be inflationary. When there might be a second currency in use that is particularly undesirable.

Taxes are inefficient.

So, most especially in a crisis when there is little chance of any of these issues being addressed development aid makes sense. This is such a situation: the world needs vaccines. There is no MMT argument against this.

So what additional tax? This is where tax gap theory comes into play. I have written on this. As I note in the linked chapter, there are five tax gaps:

1. The tax base gap, which refers to the cost of tax bases not taxed by choice e.g. wealth;
2. The tax rate gap, which refers to the costs (both positive and negative) of granting higher and lower rates of tax that vary from the norm or standard rate as well as the cost of all allowances and reliefs granted to taxpayers, for whatever the reason;
3. The cost of tax evasion;
4. The cost of tax avoidance;
5. The cost of bad debt i.e. declared sums owing but not actually paid.

These can be linked to the identity noted above for a country not relying on aid as part of its income, which is:

G = ∆B+ ∆M + T

In particular, in this equation T is tax actually collected i.e. it is a measure of cash flow. However, as the IMF has argued, this net cash sum collected is stated after two tax gaps have been deducted from total potential tax yields i.e.

where T_t is the total potential tax due on the tax base, T_f is the net tax foregone as a result of policy decisions and T_c is the tax compliance gap.

It is important to note that these expressions can, in turn, be expanded, so that in the normative typology of the tax base that the IMF suggests be used for estimation of tax policy gaps:

where T_b is the tax base for a particular tax and T_r the standard tax rate for that tax base, and:

where T_p represents the value of tax bases not taxed as a matter of policy (e.g. wealth) and T_s represents the value of allowances, reliefs, and varying tax rates granted within bases that are taxed to encourage varying taxpayer behaviours by way of tax spends, and:

where T_e is the part of the tax compliance gap resulting from tax evasion; T_a is the part resulting from tax avoidance and T_u is the part of the tax compliance gap resulting from non-payment of tax debts, or unpaid taxes.

Substituting this understanding in the equation for G:

This expanded explanation for the financing of government spending (G) offers new bases for interpretation of both the tax gap and the role of fiscal policy within the macro-prudential regulatory framework by making clear the relationship between borrowing, changes in the government created money supply and the tax gap in all its aspects.

The implication is that if there is an aim to increase revenues then there are many choices available, by no means all of which involve additional taxes from the tax bases that we are all familiar with. In this case I, and others, are suggesting the use of a tax base effectively untouched at present. That is, a tax on financial transactions (and I do of course know about the UK's stamp duty, but it has limited scope). We are putting the additional revenue claim into the first tax gap i.e. we are using a base not previously subject to claim.

Is that reasonable? The incidence of this tax can be argued. Some say because it hits pensions funds it hits those on low income who have pension investments hard: I do not agree because the answer there is to simply reduce the churn of investments within the fund, which is usually advantageous to decision making in any event.

What I do suggest is that this tax is actually beneficial in addressing another of those purposes of taxation, which is to address inequality. As I showed a little over a year ago, in the period 2011 — 18 the national income of the UK was £13.1 trillion, and in that same period the increase in net wealth was £5.1 trillion. It is stressed, that this figure is not for total wealth, but the increase in the value of that net wealth in that period. In that same period the overall effective tax rates on income during this period were unlikely to have averaged more than 29.4%, but those on wealth increases did not exceed 3.4%. In that case, if these rates had been equalised it would, at least in principle, have been possible to raise an additional £174 billion in tax revenue per annum from the owners of wealth.

Moreover, because there had been no attempt at equalisation and because the distribution of the ownership of wealth varies substantially across the UK, which variation is reinforced by factors such as age and gender where substantial inequalities exist, the effective tax rate of the 10% of those in the UK who were in the lowest-earning group of taxpayers exceeded 42% of their combined income and wealth gains in a year, but the equivalent effective tax rate for those in the highest ten per cent of UK taxpayers ranked by earnings was less than half that at just over 18 per cent. This is summarised in this chart:

As a result I suggested that there was considerable additional capacity for tax to be raised from those who own most of the wealth in the UK, many of whom are in that top ten per cent of income earners. I am now suggesting that a financial transactions tax would help achieve that goal. It would prevent ministers saying there is no capacity to tax, because there is. And the incidence would be on those with the greatest capacity to pay, whose savings have, we know, been considerably increased by quantitative easing used to address Covid related economic issues.

It would also prevent ministers from saying we must spend on other things, because we can.

It will stop ministers saying that this tax will hit ordinary people who cannot pay, because it will not.

And it will not hit borrowing or money creation if they refuse to consider either. Nor is there likely to be a spillover impact into other taxes.

The result is that this is the currently political possible way to fund what is desperately need, which is a vaccination programme for the world.

That programme is technically possible. It can be funded. The arguments against doing so are virtually impossible to find using this logic.

Will it happen? That is another question, altogether. But ministers can be held to account for their failure.