Photo: Flickr, kiwi vic.

The labour dispute at Marikana has thrown the spotlight on wages. With half of all workers paid too little to advance out of poverty, and industrial unrest rising, SA is fast running out of options, writes Claire Bisseker.

Any debate about wages in SA has to start from one crucial point: the fact that median salaries are too low - given SA's high cost of living and high dependency ratios - to allow up to 50% of workers to break out of poverty.

These are SA's working poor. On their income of up to R3000/month, most will never own a decent home, send their children to good schools, enjoy quality health care or many of the other basic ingredients necessary to unlock life's opportunities.

Their frustration runs deep and may be contributing to the increasingly violent nature of industrial unrest. At the same time, SA firms warn that they are becoming uncompetitive, partly because labour is relatively expensive compared with that in other emerging market countries.

The upshot is that efforts to diversify the economy away from its capital-intensive bias into more labour-intensive export industries have failed to make headway. Unemployment remains stuck at around 25%.

Not that creating more jobs would necessarily solve the working poor's problem.

SA's low median wage rate, and the likelihood that most new jobs will continue to be created in the low-skill, domestic-orientated services sector (where wages are low relative to the more productive export sector), means that working poverty is set to remain a central concern in SA for years, possibly even generations.

This assessment has such profound implications for wage-setting and economic policy that it is hardly surprising that SA's wage statistics are fiercely contested.

On the one hand is Stats SA's Labour Market Dynamics report. Released last month, it shows that in 2011 the median monthly earning across all sectors was R3000. Since SA workers support four to six dependants on average, this means that half of all working people are in households that skirt Stats SA's unofficial, inflation-adjusted poverty line of about R600/person a month. These are the working poor.

The problem is that this data is supplied by workers in response to questionnaires as to their gross earnings before deductions, excluding bonuses and overtime. It relies on people to be honest about how much they earn and not to confuse their gross and net earnings.

A very different picture emerges from Stats SA's latest Quarterly Employment Survey. It shows that the average monthly earning in SA's nonfarm, formal sector in May 2012 was R13647. This figure is derived from employers' payrolls. It includes bonuses and overtime payments but, because it is an average, not a median figure, it is inevitably skewed by a few very large salaries at the top end.

Economist Mike Schüssler calculates controversially (using data from the SA Revenue Service, national treasury, BankservAfrica and various other payment sources) that the median salary for full-time employees in the nonfarm, formal sector is about R10000/month.

It is worth remembering that SA's median minimum wage is R4000/month (this ranges from a measly R1503/month in agriculture to R4743/month in mining). However, compliance is extremely poor.

A study by University of Cape Town economist Haroon Bhorat et al found in 2010 that 45% of covered workers were being paid less than the legislated minimum wage. On average these workers were receiving 36% less than the legal minimum.

Steering a course between these wide-ranging estimates is difficult but, even if Stats SA's R3000 monthly median is a bit of an understatement, it is hard not to conclude that most black SA workers are earning too little to make any headway. This is because they often support large numbers of dependants, including school-leavers who themselves are unlikely to find work. In the case of migrant workers, they sometimes support two households.

"There seems to be a high-intensity conflict raging just below the surface among large parts of the working population, in many instances apparently focused on the little day-to-day money left in the hands of households after any employer deductions and financial debt payments," says First National Bank chief economist Cees Bruggemans.

He fears that the increasingly tight financial situation of many low-income households is contributing to outbursts of anger, possibly already seen in public-sector strikes in 2010, industrial strike action in 2011 and especially in mining strike action this year.

Labour Research Services (which researches labour issues for various Cosatu affiliates) argues in a recent paper that miners should be paid a "living wage" of R11702/month (close to the amount Lonmin's rock drillers have been awarded after their 22% increase).

It derives this amount by calculating that this is what would be needed for a worker to be reasonably able to afford the R2702/month bond repayment on a R300000 low-cost house.

Home ownership is most people's first big financial hurdle. Once they've crossed it, they're under much less financial stress and this means less pressure on employers to keep raising salaries. Mindful of this, mining companies have embarked on numerous projects to subsidise home ownership for their employees.

For instance, in Anglo American Platinum's Northam project, workers have been given a housing grant of R60000 each. The rest of the amount they have to finance through a normal mortgage bond.

But even though workers also get a R2500/month home-owners' allowance towards the cost of servicing a bond, only 15%-20% of the initial applicants succeeded in getting home loans, often because they held too much other debt. Many of those who were unsuccessful were put through debt rehabilitation programmes so they could reapply a year or so later.

"Household earnings need to be high enough relative to the cost of living to reasonably allow families to gain traction, stay out of crisis, and sustain an acceptable standard of living that enables human capital development," says Human Sciences Research Council research fellow Miriam Altman. "At the same time, we need industries to be competitive."

The way to make industry more competitive, however, need not be for it to pay its workers less. The consequence of that would doubtless be widespread social upheaval.

The solution, according to one of SA's foremost benchmarking analysts, BMA Analysts chairman Justin Barnes, is to make labour and management more productive.

One way to lower a firm's labour costs without lowering wages, explains Barnes, is by using labour to flex capital. For instance, if a plant that is being run on a single eight-hour shift is instead run for 16 hours over two eight-hour shifts, the firm is better able to recover its overhead costs, while maintaining its labour cost base.

Turkey's clothing industry is a case in point. It is 15 times bigger than SA's but its workers are paid far more generously. In Turkey a clothing machinist earns US$700-$800/month. In SA the metro minimum wage is less than $400/month.

According to Barnes, the crux of Turkey's competitiveness as a clothing manufacturer lies in the flexibility of its manufacturers to respond rapidly to fluctuations in market demand.

Though Turkey's working week is defined as 45 hours, employers have the flexibility to increase that to 54 hours at their own discretion. They have to give their workers only 24 hours' notice for a shift extension of up to three hours. In SA, firms can't run extended or additional shifts without incurring punitive penalties.

"Turkey shows that it is perfectly possible to compete in an industry as cost competitive as clothing with higher wages, but if you are paying labour a lot, it has to be very productive and flexible," says Barnes.

He feels that central bargaining agreements tend to limit the flexibility of firms to respond to competitive pressures and that over and above determining some national or centrally determined base wage, parties have to be able to undertake detailed plant-level bargaining if productivity is to be raised.

"In SA our arguments are so crude, so ideologically driven, that we always disagree on the basics and never get down to the detail, to the intricacies of the decisions that have to be taken," says Barnes. "As a result, SA just stays exactly where it is. We never move forward."

The initial response to the Marikana tragedy has been to hike affected workers' wages by up to 22%, with Cosatu also touting the introduction of central bargaining as a solution to labour unrest in the platinum industry. It seems inevitable that this route will lead to less job creation down the line.

SA's clothing industry, on the other hand, is moving in the opposite direction - towards discounting minimum wages coupled with greater use of plant-level, performance-linked agreements.

Though productivity growth does affect the wage-setting process in SA, the link between pay and productivity is weak - far lower than SA's peer group average.

This is the main finding of a recent International Monetary Fund (IMF) working paper by Washington-based IMF economist Nir Klein. Compared with 18 other emerging market countries, Klein found that SA was in the bottom five of the sample regarding contribution of labour productivity to real wage growth.

In fact, Klein estimates that roughly a quarter of the 750000 jobs shed by SA in 2009/2010 in the wake of the global financial crisis was because of "excess wage growth".

At a time when the domestic economy was contracting, firms were closing and jobs were being axed, SA inexplicably loosened the reins. Those lucky enough to hang on to their jobs in both the private and public sectors enjoyed rapid real wage growth far in excess of any rise in their productivity.

"This suggests that in SA, real wage growth is largely driven by other factors which delink it from labour market conditions," concludes Klein. He suggests that SA's collective bargaining framework is largely to blame.

According to a recent paper by Bhorat, being covered by central bargaining adds almost 10% to the average nonunionised worker's wage in the private sector. For a public-sector worker, the combined advantages of being unionised and covered by central bargaining confers a wage premium of about 20%.

This premium is in no way linked to worker productivity or performance since it's not as if the nonunionised workers outside central bargaining arrangements are 20% less productive than their contemporaries, just less organised and regulated.

The IMF's view is that SA's labour market arrangements have allowed entry-level wages to be set above the productivity of inexperienced workers. In fact, SA has the highest ratio of minimum wages to worker productivity among 17 other emerging market countries.

These institutional arrangements have served to protect those with jobs (insiders) at the expense of the unemployed (outsiders) and contribute to preserving inequality and unemployment in the country, says the IMF in its recent Article IV assessment of SA.

"High margins in product markets and wages in labour markets have resulted in uncompetitive domestic costs of production, eroding external competitiveness, and excluding part of the population from formal economic activity," it concludes.

The showdown in SA's clothing industry between firms that are complying with minimum wage law and those that don't is a case in point.

About 40% of all clothing workers nationally are employed in firms that are not wage-compliant. This suggests that the minimum wage agreed to between the SA Clothing & Textile Workers' Union (Sactwu) and formal-sector employers is priced beyond what the market can bear.

This is borne out by the fact that despite a 15-month grace period which ended in April this year, by far the bulk of noncompliant firms were still unable to meet the minimum wage threshold even when faced with the threat of closure.

The minimum metro wage for a machinist is R788/week. According to bargaining council figures, the worst offending firms in KwaZulu Natal pay only 56% of this or R440/week.

Last month, SA clothing manufacturers obtained a labour court order compelling the bargaining council to shut down 400 clothing firms in rural KwaZulu Natal that are failing to pay workers the minimum wage. The move will make up to 10000 workers redundant.

To the IMF and many private-sector economists, this action typifies what is wrong with SA and why it continues to fail to dent its 25% unemployment rate: the country's institutional arrangements protect "insiders" at the expense of everyone else.

Given SA's unemployment crisis, one solution would be to lower or scrap the minimum wage so as to legalise noncompliant firms and save thousands of jobs. But even if it were politically feasible, doing so in the face of SA's working poor problem would be a recipe for widespread social instability.

"The irony regarding the wage demands of workers at Marikana is that, contrary to the IMF's view, the striking workers regard themselves as outsiders," says Sanlam group economist Jac Laubscher. "This underlines the complexity of the labour-market issues brought to the fore by the Marikana events."

But why should miners earning nearly R10000/month consider themselves outsiders? Perhaps because for some their living conditions (in shacks with streets that run with raw sewage) remain those of marginalised communities.

Many are netting around R4500 a month after deductions, which is not much on which to raise an extended family. It wouldn't matter so much if the state were providing affordable housing, clean water, adequate sewerage provision, and quality education and health care.

But, according to a recent World Bank study of inequality, the provision of these basic public services is so inadequate in SA that poor township and rural communities might as well be living in Honduras, Nicaragua or El Salvador. In short, they are being denied the opportunities to rise above their circumstances.

This suggests there is an essential role for the state to play in raising what development economists refer to as the "social floor".

Among other things, this would involve strengthening the delivery in education and health services, improving access to social services and reducing the cost of living for the poor through, for example, stable food inflation, the provision of subsidised, reliable public transport and an affordable energy price path.

"The net effect would be to reduce the poverty line and to improve access to services that enable human capital development and class mobility," says Altman.

It would also reduce the pressure on employers to keep raising wages.

There is a range of other obvious things government should be doing:

• Significant labour-market reform (to align wages with productivity);
• Special interventions, like the youth wage subsidy, to create first-time employment opportunities for the youth; and
• Encouraging skilled immigration to reduce the shortage of scarce skills that is placing additional pressure on wages.

However, all these remedies are politically unpalatable to varying degrees. Given government's weak institutional capacity, it may also be too much to expect any notable improvement in the quality of service delivery any time soon.

Clearly, unless the leadership in government, business and labour are prepared to do things differently, SA is going to remain stuck in a middle-income, low-growth trap with high unemployment, a frustrated mass of working poor and rising social discontent.

If Marikana was SA's wake-up call, why is the nation still showing no sign of coming to its senses? It is beginning to look as if only "a real derailment", to quote Bruggemans, will be enough to forge a meeting of minds between employers, unions and the state.

It has taken the decimation of SA's clothing industry to inject pragmatism and compromise into its labour market. Let's hope SA doesn't have to be brought to its knees before the same spirit of give and take can take root in the rest of the economy.