/ 9 May 2020

Three ways to finance the Covid-19 policy response

It’s evident that economic inequality is rife in South Africa. Income and consumption inequalities are high and wealth inequality is even higher – much higher than global wealth inequality.
More wealth taxes may soon be a reality for east African countries in the wake of Covid-19. (Madelene Cronje/M&G)


In a previous article we called for distributive fixes to the impending humanitarian crisis facing a Covid-19-affected South Africa. As the economy contracts dramatically under the lockdown, we can expect a correspondingly dramatic decline in the government’s usual stream of tax revenue. 

Here we wish to address various ways of financing a fiscal response to Covid-19 from a pro-poor perspective. We group financing measures into three broad categories: direct taxation, de facto taxation and debt. Here, we discuss the strengths and weaknesses of each category.

Direct taxation

In the first category, we support the various calls for a solidarity tax on incomes. When working out rough estimates, it is important to recognise that it is likely that high incomes will decline as a result of lockdown-related layoffs. The Treasury projects a decline in wages and salaries of 5% to 15% over the year, depending on the severity of the lockdown. 

Given that personal income tax was projected to be R550-billion in 2020 pre-Covid19, a 10% decline in salaries roughly implies R55-billion less projected revenue. Using South African Revenue Service (Sars) tax statistics for 2019, we estimate a solidarity tax of 20% on marginal incomes of the top 4% of earners could raise up to R110- billion in additional revenue. Applied only to the top 2% of earners, the corresponding figure is nearly R80-billion.

We believe there are strong moral and, importantly, economic grounds for expanding personal income taxes on high-income earners during the current moment to finance the necessary state expenditure on low-income households. The first reason is that a disproportionately high percentage of the rich are able to earn income, just as they usually would, by working from home, as noted by academics Amy Kerr and Andrew Thornton, while low wage-earners are more likely to suffer lay-offs. Policy should correct the regressive effects of this feature of the lockdown.

The second reason is that the consumption of high earners who retain their salaries is significantly restricted by lockdown regulations. This seems to indicate that a significant portion of rich households are adding considerable amounts in savings to their stock of wealth during the current crisis. 

Using data from the Income and Expenditure Survey (IES) of 2011 and assuming the consumption patterns by top percentiles are stable, we estimate that the top 4% of salaried workers annually spend about R40-billion on alcohol and restaurants, R100-billion on transport, and R40-billion on durables such as clothing and furniture. Under level-five lockdown regulations, all of these items are severely restricted. Under level one, restrictions on consumption are lower but still substantial. 

How much finance could a solidarity tax raise?

Tax threshold (earnings per year)350 000500 000
Number of people1 427 932757 396
Percentage of adults4%2%
Total taxable income 1 192 000 000 000870 000 000 000
Taxable income above tax threshold611 000 000 000430 000 000 000
Projected income accounting for lockdown550 000 000 000387 000 000 000
Revenue from 20% tax increase 110 000 000 000 77 000 000 000
Savings from reduced consumption  
Upper bound: Level-five lockdown181 000 000 000111 000 000 000 
Lower bound: Level-one lockdown60 000 000 00036 000 000 000
Notes: Authors’ calculations. Tax data from Sars Tax Statistics (2019). Consumption estimates from IES (2011), where level one includes expenditure on alcohol, restaurants and vacations, as well as 10% of transport; and level five additionally includes expenditure on durables and all transport. 

Although these are rough estimates, because it is difficult to predict the impact of the lockdown on spending patterns, there is a strong case that there are substantial increases to the saving of rich households. Taxing accumulated savings could be done without some of the usual negative repercussions present under recessions. 

Usually one might be apprehensive about raising taxes during a downturn for fears of stifling aggregate demand. These fears may not apply in full to well-targeted taxation during the Covid-19 lockdown: the lockdown itself imposes the restrictions on non-essential consumption, meaning increased taxation on top-income earners might not generate harmful economic effects on the economy through depressed aggregate consumption. 

Essential consumption is unlikely to diminish greatly with a substantial tax on top incomes (if your after-tax weekly income halves from R25 000 to R12 500 it’s unlikely you’ll move from consuming two loaves of bread a week to just one) and non-essential consumption such as clothing, furniture or transport seems unlikely to surge after the lockdown. 

In short, savings during the lockdown are surely increasing for a subset of the population and it is unclear that we should treat these accumulated savings as pent-up demand waiting to explode the moment lockdown ends. In that light, a just (and potentially expansionary) use of these accumulated savings can be to redistribute it to those who need and will use it.

Other direct taxes such as a wealth tax, as recently argued for by Aroop Chatterjee, Amory Gethin and Léo Czajka, might be similarly used; even if the infrastructure for such a tax takes time to create, it will surely yield benefits over a longer period. It should be acknowledged here that proposals to increase some forms of taxes is out of keeping with the international policy response. Analysis of the policy response of G20 countries suggests that in aggregate these countries have cut taxes to the tune of just less than 1% of gross domestic product (GDP).

De facto taxation

The second category involves price controls to ensure food is available to households at accessible prices. Using the 2011 IES data, the bottom half of people in terms of income distribution (who are also under the national poverty line) use about a third of their monthly consumption-spending on food: 80% of this food expenditure is from either chain or other retail stores. 

The prices of these essential goods, therefore, have an immediate impact on household food security. Price controls are currently implemented on a narrow range of goods: this range should be monitored and potentially expanded to include other basic goods such as bread, sugar beans and eggs.

If the state were to impose a strict control on the mark-ups of necessary goods (or a subset of these goods), it leaves space for firms to leave mark-ups on non-price-controlled goods (let’s call these goods “luxuries”) unchanged, or to raise mark-ups on these goods as a means for offsetting the negative effects of the price controls on their profit margins. In the case of the former, the price control acts as a corporate tax and in the case of the latter it acts as an income tax. 

Both of these taxes are likely to be progressive interventions if the goods defined as luxuries fall disproportionately into the consumption baskets of higher-income groups. Price controls on necessary items also guard against retail companies capturing an undue share of the grant expansion the government has committed to. 

Debt-financed policy

The possibilities for financing the expenditures rendered necessary by the Covid-19 shock extend beyond raising tax revenue to our third category, debt-financed policy. In this regard the government has three main options. It can sell bonds on the market to domestic or foreign private actors; access credit facilities at international financial institutions, such as the New Development Bank and International Monetary Fund; or sell its bonds to another branch of the state, such as the Reserve Bank. There are advantages and disadvantages to each avenue that require consideration. 

The first option raises concerns about the stability of the public-debt-to-GDP ratio, considering the high cost of borrowing in a moment when demand for public-sector debt in emerging markets is low. High interest payments on public debt could come to squeeze out expenditure on socially necessary budgetary components in the future. 

The second option raises concerns for analysts who worry about the country’s policy autonomy in the light of institutions such as the IMF’s pernicious history of structural adjustment. It should be acknowledged here that emergency financing might not come with the conditionality assumed by many critics: if the government borrows, say, $5-billion from an international finance institution, it is unlikely that this institution can, as a condition of that loan, demand the privatisation of a domestic public institution worth $50-billion. Loans from these institutions can also carry the benefits of coming in “hard currency” and at concessionary interest rates.

Critics of the second strategy argue that all finance should be raised domestically and in domestic currency to diminish exchange-rate risk. They thus argue for the third strategy. The third strategy has the benefit of enabling the state to set borrowing costs at nil. However, advocates for exclusively using this financing approach fail to address concerns about the need for hard currency. Unless the South African Reserve Bank purchases a counterfeiting machine to print dollars or breaks into a Spanish bank and prints its requirement of euros, à la Netflix drama Money Heist, it cannot originate the country’s hard currency requirements. 

As such, this strategy leaves South Africa vulnerable to its need for a host of imported goods, including the urgent medical supplies required during a pandemic. Furthermore, even if we assume that there is substantial excess capacity even while the government restricts output during the lockdown, while monetary-financed stimulus might not generate inflation through generating excess demand, it can generate inflation through the external channel, if an announcement of monetary-financed fiscal expansions generate self-fulfilling expectations of depreciation. In short, advocates of monetary stimulus must engage with the ramifications of the subordinate position emerging-market economies find themselves within the global economic system.

In conclusion, as argued above, there may be low-hanging fruit for financing the economic policy response through well-targeted taxation, as the rich accumulate savings during the lockdown. However, in the light of the depth of the economic downturn, and to avoid a pro-cyclical policy intervention, new taxes cannot be the only financing means. 

There will have to be a role for price controls as a de facto tax as well as new debt, including perhaps a role for direct purchasing of Treasury bonds by the Reserve Bank. As such, nuance might dictate some kind of blend of financing approaches, in recognition of a role for tax financing to deliver distributively just outcomes; central bank financing to minimise growing interest payments crowding out social expenditure (while considering trade-offs in the form of further rand depreciation), and lending from international finance institutions to deliver the country’s hard currency requirements. 

Adam Aboobaker and Ihsaan Bassier are PhD students in the economics department at the University of Massachusetts, Amherst