Picture: 123RF/Natdanai Pankong
Picture: 123RF/Natdanai Pankong

When the government hiked VAT last year, after many years of personal income tax increases, it did so on the basis of economic efficiency — because studies showed that, of the major taxes the government could potentially raise, increasing VAT would cause the least harm to growth and job creation.

However, a new study has found there is a welfare cost to hiking VAT, not at the aggregate level, but at the level of the poorest of the poor — and that this group can never be fully compensated for VAT hikes by increases to social grants, because they mostly don’t qualify for these benefits.

The study, by a team of local and international academics collaborating under the Southern Africa — Towards Inclusive Economic Development (SA-TIED) banner, found that it would be more effective to alleviate poverty by scrapping the zero-rating of 19 basic food items and instead use the revenue generated — almost R20bn a year — to target a new social grant at the poor who currently fall outside SA’s social security net.

This is a controversial suggestion in SA, where the labour movement and many civil society organisations have lobbied hard for the extension of zero-rating.

Last year, acting under intense public pressure, the National Treasury added sanitary towels, white flour and cake flour to the list of zero-rated items on the recommendation of an independent advisory panel. These additional items were legislated during parliamentary hearings last year, and signed into law this January.

SA-TIED is a collaboration between the UN University World Institute for Development Economics Research, SA’s National Treasury, the department of trade & industry and many other governmental and research organisations in SA. Though the SA-TIED research team included people from the SA Revenue Service and the Treasury, their working paper on VAT does not reflect the views or policy positions of these institutions or the government, but only of the authors.

The paper finds that though the increase in VAT from 14% to 15% in April 2018 was coupled with an increase in social grants, it still caused a slight rise in poverty and inequality. This was reversed in October 2018, when the government increased social grants for a second time as a specific remedial measure.

But even after the second increase, the poorest 10% of the population experienced a 5% reduction in their purchasing power as a result of the VAT hike. This is because many in this group do not qualify for social grants — they are not disabled and are aged between 18 and 59, which makes them too old for the child support grant and too young for the old age grant.

Among the poorest 50% of the population, just over 30% of households receive no grants; among the poorest 70%, the figure is 40%, according to the independent panel’s report.

To find ways to reduce the impact on these households, the researchers modelled various reform scenarios in which the VAT hike was retained but zero-rating abolished, and the revenue generated was ploughed back into the social grants system.

Three new youth grants of R200, R225 and R280 a month were modelled, targeting those aged 18-30 years. Three grants for older people were also explored, introducing a new benefit of R120 or R200 a month for those aged 26-59 years, and R225 a month for those aged 36-59 years.

In all six cases, the beneficiaries were either means-tested or had no employment or self-employment income, and were not receiving the disability grant.

All six of the reform scenarios reduced SA’s overall poverty head count ratio and inequality slightly. All of them also increased the purchasing power of the poorest decile, reversing the impact of eliminating zero-rating.

"Thus, even if the 1% VAT hike is retained, the detrimental impact on the poorest decile can be reversed by rerouting revenue forgone through zero-rating into new social benefits," the researchers conclude.

The most effective scenario of the six was the introduction of a R200 monthly benefit for unemployed 26-to 59-year-olds.

It raised the lowest decile’s purchasing power by 41%.

The researchers were also interested in establishing whether scrapping zero-rating would allow the extension of a universal grant to all SA adults.

Two universal benefits of R200 a month were tested for different age groups: all adults aged 18-59 years and those aged 18-30 years. Though the study found that a universal grant would do more to reduce poverty than the means-tested benefits, the annual cost would not be revenue neutral (it would far exceed R20bn), even if the grant paid just R200 a month.

A R200 universal grant to those aged 18-59 years would reduce SA’s poverty head count from 33.5% to 29.6%, but cost an extra R51.3bn a year, over and above the revenue recouped from scrapping zero-rating. If restricted to 18-to 30-year-olds, a R200 universal grant would reduce poverty to 31.7%, but about R30bn extra would need to be found.

In the final scenario, the VAT hike was reversed, zero-rating eliminated and the revenue generated used to increase the child support grant by R40 to R450 a child per month. Though this reduced poverty from 33.5% to 32.8%, it failed to provide significant redress for the poorest.

Taken together, the simulation results suggest that a policy package of a uniform VAT rate and an expanded set of social benefits would lead to reduced poverty and inequality, in comparison with the current practice of zero-rating some consumption goods.

It also demonstrates the "poverty-zapping power" of government resources when they are directed through the social grant system rather than indirectly through zero-rating, substantiating arguments that have been made in several other local and international studies.

The paper vindicates the Davis tax committee (DTC), which recommended in 2015 that no further items be zero-rated.

It concluded that zero-rating is "an extremely blunt and second-best instrument" for addressing equity considerations, and that redress could be better achieved through the expenditure side of the budget.

Just like the SA-TIED researchers, Davis concluded that it would be better to collect the tax revenue and redistribute the additional income through a targeted transfer to the poor. However, the DTC conceded that it would likely be difficult to eliminate the current zero-ratings entirely.

The National Treasury discovered just how hard that would be, when it went before parliament to defend the VAT hike early this year.

"What we argued in parliament is that we’ve got too many zero-rated items, and many of these are not well targeted to poor households," says the National Treasury’s head of tax and financial sector policy, Ismail Momoniat. "But we were finally required to give more zero-rated items. So, whatever an academic paper says, and we cannot rely only on one paper, it’s not always possible to achieve the pure research position, as we still have to hear the views of other key stakeholders who provide comments to draft tax proposals. There are lots of other factors to take into account before any policy is finalised, including the capability of the state in targeting poorer households.

"If it were based purely on academic research, and the state had good capability, we’d be able to abolish zero-rating and replace [it] with better targeted grants [for] poor households," he adds. "What you end up with in reality is a combination of the pure research, the capacity of the state, and submissions from labour, civil society, business and parliament."

One limitation of the paper is that while it outlines the positive welfare benefit of introducing a new social grant, capped at a revenue-neutral R20bn a year, it doesn’t consider the cumulative fiscal burden of introducing a permanent new entitlement in a climate of high unemployment and low growth.

Over the past decade, the number of social grant beneficiaries has grown by 44%, from 12-million in 2007 to 17.2-million now, largely due to the expansion of the child support grant.

At present, about one in three South Africans receives a cash transfer. According to World Bank data, SA has the second-largest share of households receiving state transfers in the world, after Iran.

Given SA’s current fiscal constraints and long-term fiscal risk profile, it is abundantly clear that there is no fiscal space in which to introduce any new social grants, even if initially funded by scrapping zero-rating.

Further VAT hikes are also presumably off the table. The public backlash against the 2018 VAT hike, coupled with the key finding of this paper — that the poorest suffer the most from VAT hikes and can never be fully compensated through the existing social grant system — will make it doubly difficult for the government, should it seek to hike VAT again.