Despite widespread calls that data prices must fall to broaden internet access, government and business remain at loggerheads on the subject.
Telecommunication has been earmarked as a sector that can promote inclusive economic growth, but reforms have been slow and marred by controversy.
A World Bank study has shown that for every 10% rise in broadband penetration, GDP growth increases by 1.38%.
It is against this background that there have been calls to zero-rate the price of data, which currently attracts VAT at 15%.
National Treasury recently broadened the terms of reference of the independent panel tasked to review the list of zero-rated items, and it may now consider submissions on new “non-food” items. The current zero-ratings basket only includes 19 food items. The panel has to finalise its report by the end of July.
But industry insiders believe calls to zero-rate data are unlikely to find favour.
Gerhard Badenhorst, director for tax and exchange control at Cliffe Dekker Hofmeyr, says there are generally three justifications for zero-rating or exempting certain goods or items:
- To reduce or eliminate the regressivity of VAT, or to improve progressivity (the current zero-rating of basic food items falls into this category);
- Merit goods or services, in other words goods and services deemed to be in the public interest (the current exemption of housing, public transport and education fall into this category);
- Goods or services that are difficult to tax (the current exemption of financial services falls into this category).
He says progressivity will only improve if the zero-rating benefits poor households substantially more than higher income households.
“I would think that data is used or consumed substantially more by high income households compared to poor households. In addition, businesses are substantial consumers of data. The zero-rating of data will therefore not in my view improve progressivity.”
With regards to the public interest, Badenhorst says if the perception is true that individuals use data mostly for social media, it is difficult to see how the zero-rating of data can be considered a merit item.
The biggest issues with data costs seem to be the expiry of unused data and out-of-bundle data costs.
“I would think that as a ‘merit’ item, medical services would rank much higher than data.”
As data is not difficult to tax, it does not fall into the third category.
Moreover, the zero-rating of data seems to fall outside the panel’s mandate. The revised mandate includes “the identification of any items other than the current zero-rated food items that will achieve the policy intention of providing relief to the poor and low-income households with particular consideration to the needs of children, women and other vulnerable groups”.
Badenhorst says the mandate of the panel therefore seems to be limited to items improving progressivity and does not extend to the identification of “merit” items. The zero-rating of data does not seem to benefit low-income households proportionally more than high income households, and it will not relieve poverty or inequality.
The Davis Tax Committee also seems to be of the view that it is preferable to rather collect the tax revenue and to redistribute the additional income to poor households in a targeted manner.
Erika de Villiers, head of tax policy at the South African Institute of Tax Professionals, says it simply does not make sense that the current pressure on data providers to reduce the cost of data should – at least to an extent – be absorbed by government not collecting VAT on data.
“Another consideration for the panel will be the amount of data used by higher income earners versus lower income earners as the panel has to take into account the absolute and proportional benefit likely to accrue to low-income households,” she says.
Dr Seán Muller, senior lecturer in the School of Economics at the University of Johannesburg, says there are a number of factors that need to be taken into account in determining whether zero-rating data would be desirable.
First, is zero-rating likely to reduce the consumer’s cost of data or will the firms selling it absorb most of the benefits? Second, even if the removal of VAT from this product is partially passed on to data prices, who will be the main beneficiaries of that?
“If high data prices are partly a function of a lack of competition in this sector, then producers are more likely to be able to absorb the benefits rather than pass them on to consumers.”
In terms of who benefits, it is reasonable to assume that wealthier South Africans buy more data and therefore in absolute terms they will benefit more (per person) from zero-rating than poorer South Africans, he says.
“The bottom line is that zero-rating is a very crude, and possibly ineffective, way of making data more accessible to poorer citizens.”
There are also concerns about South Africa’s tight fiscal position and whether such a proposal – even if viable – will be affordable.
Zero-rating data for all users will not be economically efficient as the government will forfeit the VAT revenue from both the rich and the poor in its attempt to assist the poor, De Villiers says.
“I very much doubt whether this will be affordable,” she says.
Nazmeera Moola, co-head of fixed income at Investec Asset Management, says an expansion of the number of zero-rated items will have a direct impact on revenue. Where items are widely consumed – like chicken for example – the impact can be significant.
“I can only imagine that data would be a massive impact,” she says.
Moola says given the current revenue shortage and the country’s weak fiscal position, the capacity to do any significant expansion of zero-rated items is quite limited. However, the public-sector wage settlement suggests that political imperatives sometimes triumph over fiscal sense.
Against this background, it is difficult to say what the likelihood is that data could be zero-rated, she says.
“However, if the question was ‘Can we afford to do that?’, my answer would be no.”
Muller says whatever the actual loss of revenue to the state – which would require more detailed analysis – the case for zero-rating data seems quite weak.
“Pursuing better regulation of, and competition, among data providers may be a better option,” he notes.