Reducing inequality to promote growth: a proposed policy package

Reads 35,781
Kuben Naidoo, 30 January 2013
Any growth strategy for South Africa should include elements that address inequality explicitly. This article identifies reforms that are likely to support growth in the long term and proposes a policy framework to ensure a more equitable distribution of the dividends of economic growth. These relate to high-quality education for the poor, progressive taxation, a social safety net, anti-monopoly policies and labour market reforms to promote the employment of low-skilled people.


There is considerable theoretical and empirical evidence to support the assertion that inequality is bad for growth. In unequal societies it is much harder to develop the institutions, norms, mores and conventions required for economic growth. In addition, social mobility slows to a crawl in unequal societies. This undermines the incentives for hard work and effort and weakens the human potential of a country. Inequality also damages human capital formation in ways that are hard to repair.

It follows that any growth strategy should include elements that address inequality explicitly. Some such measures may, however, impede growth in both the short and in the long run. Such measures should be avoided in favour of measures that contribute to building the institutions and creating the conditions necessary for higher growth  in the long run.

My thoughts on what should and could be done about inequality in this regard amount to five policy elements that need to be instituted as a package:

  • High-quality education for the poor
  • Progressive taxation
  • A social safety net
  • Anti-monopoly policies
  • Labour market reforms to promote the employment of low-skilled people

This article does not tackle the priority of how to raise economic growth. It identifies reforms that are likely to support growth in the long term and proposes a policy framework to ensure a more equitable distribution of the dividends of economic growth.

What is not on the list and why

Many people would want to add points to the list. The first – and perhaps the most obvious one – is higher economic growth. They argue that some sort of Kuznets curve exists and that, with higher growth, countries will reach the point of declining inequality quicker; that growth is critical to raising incomes and living standards as well as fixed investment.

Economic growth per se has been left off this list deliberately because growth can lead to either rising or falling inequality depending on the circumstances and the nature of that growth. China’s rapid economic growth has coincided with rising inequality while Brazil’s growth acceleration has led to declining inequality. Angola’s resource boom has led to rising inequality while Vietnam’s agriculture boom has led to falling inequality. Within the rich world, some countries such as the US and the UK have become more unequal with growth while Germany has become more equal. Growth on its own provides no guarantee of declining inequality.

The second factor that people may want on the list is a competent, effective government. It is argued that progressive taxation makes sense, and that tax morality can be sustained, only if the money is spent wisely. To some extent this is obvious. However, the empirical evidence does not support the assertion unequivocally. The UK has one of the most capable bureaucracies in the world, but inequality has risen steadily. Brazil has quite a poor track record in delivering education or safety in poor neighbourhoods, yet inequality has fallen. The three countries in the world with the fastest rise in inequality (the US, UK and China) all have fairly competent governments.

A third candidate for the list is corruption. People argue that corruption hurts the poor the most and gives disproportionate power to citizens with money. However, the empirical evidence suggests that the causation runs the other way: as countries become more equal, fighting corruption becomes easier. A growing middle class often is associated with falling corruption. 

What is on the list and why

The reasons for listing good quality education for the poor are self-evident. In a world where the rates of return to education are rising rapidly, good quality education for the poor is the surest way to increase social mobility and to ensure that workers can benefit from improvements in their productivity.

Good quality education for the poor is hard to achieve – especially in an unequal society. Unequal societies tend to reproduce inequality through the education system. To counter this, concerted action is needed to promote quality education in poor communities. It will cost much more to produce similar outcomes in poor schools than it does in richer communities. A series of institutional reforms could be undertaken to improve the quality of schools in poor communities. Most of these are beyond the scope of this paper, but more local accountability for school performance has to be part of the equation. Better quality in post-school education and training is also critical.

To tackle inequality, we also need a progressive tax system. The more unequal the country, the more steeply progressive the income tax system should be. There are several reasons for implementing steeply progressive taxation. The first is that some of the income that accrues to the rich is from economic rents, not effort. These rents include ownership or senior employment in monopoly or oligopolistic industries, inheritance and capital gains.

Secondly, it is true that taxation can make a rich person poor but cannot make a poor person rich. However, progressive taxation is an effective instrument to reduce inequality and thereby promote economic growth because less inequality makes it easier to build respect for the instructions critical to growth (such as the rule of law and mechanisms for the non-violent resolution of disputes).

There are both technical and political-economic challenges in making the tax system more progressive. Rich people genuinely believe that they earn what they earn because of skill, hard work and effort. This is not always the case, especially given our apartheid history and, more recently, elements of the BBBEE and tenderpreneur experiences. Nevertheless, there are risks in making the tax system more progressive, notably an exponential rise in the incentive to avoid and evade taxes.

A strong social security net is necessary both to protect the vulnerable against life’s uncertainties and to make the labour market more efficient. Elements of such a safety net are unemployment insurance, disability benefits, retirement benefits and health insurance. This is one of the few areas where South Africa has done well, though gaps remain. Unemployment insurance should be broadened and contributory retirement benefits for low-income workers introduced.

In addition to ensuring that no one lives below a certain living standard, a well-designed social safety net enables the labour market to be more flexible and efficient. Employees can freely look for opportunities and employers can get rid of non-performers at a lower cost to the firm. The risks and vulnerability of poor work seekers in particular are reduced. Efficient labour markets are critical to ensure economic dynamism, i.e. the ability of people to move out of and into firms and sectors.

Anti-monopoly policies are essential to reduce economic concentration over the longer term. At least part of the reason for high levels of inequality is that monopolies extract large rents from the economy, distort capital markets and inhibit innovation and entrepreneurship in the economy. Exactly what to do about monopolies and oligopolies is less obvious. A combination of tougher competition laws, deregulation in certain parts of the economy and tougher regulation of monopolies in other parts of the economy is needed. In addition, trade policies and investment policies, government procurement and taxation should be used to stimulate competition. There are still several areas, most notably in the telecoms sphere, where barriers to entry are state-imposed, at the expense of investment and productivity in the wider economy.

Labour market reforms

A last requirement is for labour market reforms aimed at promoting the employment of people with low or medium skills - a controversial proposal in the South African context. A major cause of the high levels of inequality in South Africa is that too few people work. If many more people were employed, even in relatively low-paying and low-productivity jobs, inequality would fall. South Africa needs to create millions of jobs for the presently unemployed. Once people have entered the workplace, they can improve their livelihoods through growth in their productivity and resultant higher earnings.

Given that the vast majority of unemployed people lack appropriate skills, major reforms are required to support the entry of such low-skilled people into (probably) low-productivity jobs. Among the reasons why South Africa employs so few people in low-skilled jobs are high entry-level wages and numerous policies and incentives that skew the economy towards higher-skilled, higher-productivity jobs.

Much more needs to be done to enable the entry of unskilled people into low-productivity jobs.  These include reforming the labour market to match wages to productivity and to reduce the risk of hiring workers. Other policy initiatives include support for labour-intensive industries, for example subsidies, special zones, public works programmes, internships and apprenticeships. While some of these are in place, generally they are too limited to make an impact on inequality.

Reforming the labour market is steeped in political-economic challenges. There are several ways to overcome these challenges. The first is a simple deal or social compact: a stronger social safety net and higher taxes on the rich (to fund the higher education spending on the poor, the social wage and subsidies to labour intensive industries) in return for more flexibility in the labour market. A second option is to declare certain zones (in poor provinces or even in urban townships) where certain labour laws would not apply. A third option is to use age as a criterion to have a greater degree of dualism in the labour market, with existing workers enjoying benefits and protections that newer employees would be denied for a period of time.


As part of any growth strategy, policymakers in South Africa have to confront the level of inequality which stunts growth, threatens stability and undermines long-term investment. If the majority of the population perceive the system to be unjust, the system will not be sustainable or likely to be dynamic and growing.

South Africa needs to adopt a package of policies aimed at tackling inequality. Any element on its own will not do the trick. Together the package of five proposals is intended to reduce the distorting effects that inequality has on economic dynamism and to raise social mobility through employment and rising productivity. Tackling inequality sensibly, as proposed, will help to boost growth and to change the incentives for long-term investment.

Download article

PDF icon Naidoo Jan 2013 Reducing inequality for growth FINAL.pdf
Post a commentary

This comment facility is intended for considered commentaries to stimulate substantive debate. Comments may be screened by an editor before they appear online. To comment one must be registered and logged in.

I find this article interesting, however I will hazard to provide some comments, and I will avoid being technical.
Each country is a country "and South Africa is different". I heard the latter a number of times since I came to SA in 2007. But this is obvious: Wales is different from Scotland, and Norway differs from Sweden.
Brazil is mentioned by Kuben, it is very unequal, less than it used to be, but still unequal. I am not sure inequality in Brazil is being reduced because of growth. Maybe it is. However, some would say that Bolsa Familia plays a role (if you are labour economist).
It seems to me that inequality in Brazil begun its decent in 1994, when hyperinflation came to an end. High inflation was a feature in Brazil for at least two decades. Also, the old import-substitution model came to an end - a bit earlier, around 1991. By 1995 the policy framework in Brazil was very different than in 1989, when the first democratically-elected President came into power.
I will risk to say that macro-economic stability combined with better institutions and policies (e.g., fiscal responsibility laws, a certain degree of central bank independence, very flexible labour markets, economic openness ... this list can go on and on) plus a very entrepeneurial population, created the preconditions for growth to take off. With growth, yes, inequality tends to come down, that why we call it growth. All in all, in the Brazilian case, high inequality was no impediment for growth.
We have some of those above-mentioned things in SA, but not others, and perhaps we should start thinking about what is not in place here. (Having said that, it is not easy to engineer growth). For more on the Brazilian case, perhaps the following link can help,

This comment facility is intended for considered commentaries to stimulate substantive debate. Comments may be screened by an editor before they appear online. Please view "Submitting a commentary" for more information.