It is increasingly difficult for South Africans to get small mortgages and therefore own low-cost homes, despite several government initiatives. That is damaging low-income consumers’ financial health and frustrating access to housing in urban areas. We need new ideas on how to fix the problem. I want to propose one.

But first, consider the scale of the issue. Today, 66% of South Africans are urban, up from 62% in 2009. In Johannesburg alone, 3,000 new residents are thought to enter the city per month. The demand for mortgage finance should, therefore, be higher than ever. But in the last quarter of 2018, just R1.4bn was lent in mortgages of R350,000 or less, only 3.4% of all mortgage lending. Ten years ago, 16% of all mortgage finance went to that segment and R4.4bn was distributed.

There are various reasons for the collapse of small mortgages. Banks are under pressure from regulators to shorten the length of their loan books and thereby reduce risk, so are less keen on 20-year assets. But the market has also shifted into unsecured loans because there are bigger profits to be made. Anecdotally, low-income consumers are using unsecured loans as part of their efforts to buy properties (though some also do so through lease-to-buy deals). That, naturally, comes at a far higher cost. Since the National Credit Act opened the market for unsecured loans in mid-2006, growth has been substantial. At the end of 2008, 12% of the credit disbursed in the market was unsecured credit. A decade later it was 22%.

The margins on mortgages are a fraction of those on unsecured loans, plus the costs of mortgages have gone up due to court decisions that affect banks’ ability to quickly sell properties in possession. Simultaneously, the cost of new housing has risen.

This is a disaster for long-term urban stability. We want urbanisation to accompany widespread property ownership to ensure residents have a vested interest in urban development. While we have made much of the importance of giving people title to their properties in the land-reform debate, we’ve made little of enabling people to acquire title in the first place. We need interventions in the market that will shift lending patterns back into mortgages.

There are several government and city-level schemes that help. The Finance-Linked Individual Subsidy Programme (FLISP) allows households to claim a grant from government for their first homes if they earn less than R22,000 per month. The subsidy is up to R121,626 for those earning between R3,501 and R3,700, and less for those earning more. But in the 2017/2018 year, only 2,295 subsidies were granted, while over 17,000 had been budgeted for. The scheme was overhauled in July last year, increasing subsidies and qualifying limits. R950m has been budgeted for over the next three years. But even if it reaches targets, it is still a relatively small dent in the problem. The programme is administered by the National Housing Finance Corporation, a quiet but well-run state-owned enterprise. Over the past year it leveraged R2bn in private-sector funds into home financing while disbursing R381m directly itself. This is admirable, but it needs to be at a larger scale to make a noticeable difference, considering that the total SA mortgage book is now close to R1-trillion. Steps are being taken to scale it with the creation of a human settlements development bank, depending on how National Treasury capitalises it.

But clearly more is needed. So here is an idea. Tax-free savings accounts have been introduced to drive a better savings culture in SA, but further incentives could support savings specifically for housing. A scheme could allow those who earn within the lowest tax bracket of up to R195,850 per year to contribute up to R50,000 per year to a first-time buyers savings account.

Rather than a deduction, such as for pension savings, government could add 18% to amounts contributed into a first-time buyer’s account (perhaps sliding lower the more that is contributed). It would effectively be giving back the 18% that taxpayers in the lowest bracket pay, though those earning below the tax-free threshold of R79,000 would also benefit and therefore gain an incentive to save too. Providing a credit has a better psychological effect than avoiding tax through a deduction and spurs greater behavioural change.

Such a policy would be highly progressive and redistribute income into productive asset accumulation by lower-income South Africans, reversing some the inequality currently being created by differing costs of finance at different income levels. The problem with FLISP is that it doesn’t crowd in personal savings in the same way. In time such a scheme could integrate or replace FLISP.

Catalysing the credit market to direct more finance into low-cost housing is clearly needed, and ideas on how to do so should be part of the land reform debate.

• Theobald is chairman of Intellidex.