POLITICS

Why a Youth Wage Subsidy won't work - COSATU

Federation's detailed response to National Treasury and Democratic Alliance

The Youth Wage Subsidy in South Africa

Response of the Congress of South African Trade Unions

(To the National Treasury and the Democratic Alliance)

Executive Summary

The National Treasury document relies on international studies to justify its proposals on youth wage subsidies, or even employment subsidies. But careful reading of those studies reveals that they lean towards the COSATU position. Based on Treasury's own literature which they used to propose this ill-conceived policy, we have shown that there are no empirical grounds for this policy.

The irrelevance and the likely wasteful effects of this proposal have been amply demonstrated in many cases. International literature shows that the wage subsidy idea is extremely costly and wasteful, with massive deadweight losses. The ILO (2011) reports that "research in various countries has shown that wage subsidies lead to combined deadweight and substitution effects of the order of 70-90% of the number of jobs created". The estimate by National Treasury puts the deadweight loss alone to be 58%, i.e. 58% of the promised jobs from the subsidy would have been created without the subsidy (that is, if we believe National Treasury's estimates!).

The youth wage subsidy will have significant substitution effects. Firms will have an incentive to let go of existing workers in order to employ subsidised ones. The National Treasury document dismisses this concern on the basis of extremely weak arguments. National Treasury pretends as if it does not operate in South Africa, where the elementary rights of workers are violated on a daily basis.

For example, the vast majority of workers do not enjoy the minimum wage regulation. But also more pertinent is the fact that only 29% of the workforce is unionised in South Africa, which opens up the rest, 71%, to abuse. In addition, the existence of labour brokers who screen and manage workers for employers also makes it easy to fire existing workers and get "good ones" on a subsidised basis.

The substitution effects are likely to be widespread, especially with the existence of labour brokers. This substitution will hit the unskilled and semi-skilled parts of the workforce the most. By this criterion, and using the 2011 industry structure tables, we find that at least 3.7 million workers are vulnerable to substitution in the South African economy. The tables also include a category of mid-level skilled workers, whose skills composition is difficult to ascertain.

However, we can interpret the figure of 3.7 million as a minimum number of workers who are vulnerable to substitution. The National Treasury document argues that the subsidy is introduced at the beginning of a recovery, so that substitution effects will be limited. But this argument is clearly not structural as it does not consider the skills composition of the sectors involved. Secondly, even if we concede that economic growth will limit substitution, what will happen is that deadweight costs would increase, because with economic growth would lessen the potential impact of the subsidy on firms' employment.

But we also know that economic growth is not necessarily going to rapidly increase job-creation because employers do not adequately reinvest the profits generated back into the economy and the structure of the economy remains problematic. Today employers are sitting on R600 billion which they are refusing to invest in the productive economy.

The third reason why we oppose the youth wage subsidy is that it does not guarantee that training and skills development will take place in the workplace, less so in the sectors where job-creation is likely to be created: wholesale and retail trade, personal services and construction. As we have noted from the Commission on Equity and Empowerment, little training is dedicated to black people, less so to the skilled segment of the workforce.

The situation is worse for the unskilled, who are likely to be outsourced, casualised and employed through labour brokers. What is even worse, the National Treasury does not want to mandate training: "the design of any potential employment subsidy may not want to mandate training alongside the subsidy since additional administrative burdens on employers may discourage take-up of the subsidy". This is indeed a problem, because Treasury's own literature says that these subsidies depend on training for success.

As van Reenen (2003) writes: "The success of the employment subsidy option will also hinge on the extent to which the experience of work and training will raise productivity, thereby enabling workers to keep their jobs when the subsidy runs out". Without mandating training, which is currently very minimal, it is clear that National Treasury relies purely on the philanthropy of the private sector, an expectation which runs against the daily experience of the vast majority in the South African workforce.

The fourth reason why we oppose the youth wage subsidy is that it will lead to the recycling of young people without training. In the literature they say young people will be fired once the subsidy ends. National Treasury (and the Democratic Alliance) dismiss this on the grounds that "it's lousy business to fire good workers". But the fact that businesses have moved drastically towards outsourcing, labour broking and casualisation (of good workers) shows that it is good business to have a workforce that is vulnerable and flexible.

The goodness of the worker is subordinate to the power profit. Indeed there are many good workers that have been retrenched only to be hired under labour brokers, or as casuals. Because National Treasury does not mandate that business be held accountable for "recycling", and seeks to ensure that, during the subsidy period, these young people do not have recourse to labour protection; the proposed youth wage subsidy will produce massive "destructive churning".

The fifth reason why we oppose the youth wage subsidy is that with major substitution and increased vulnerability of the workforce, there will be downward pressure on wages. Inequality will worsen as low wage workers replace those that have managed to capture non-wage benefits in their compensation. It can be shown that the increase in the mark-up due to the subsidy will raise the profit-share at the expense of the labour share. This therefore will not take us forward with the triple challenges. Indeed jobs have been created, but at the level of poverty wages. In additionpoverty is likely to rise, because employed workers with relatively higher wages will be replaced by many vulnerable low wage workers.

The sixth reason why we oppose the subsidy is that there is an underlying assumption that there is a gap between entry-level wages and productivity among young workers. Treasury and the Democratic Alliance argue that youth wage rates are too high. However the National Treasury document fails to compute this gap between the wage and productivity. With an average wage of R940for those that fit the characteristics of at least 60% of the unemployed, it would be interesting to know what is expensive about this average monthly wage.

In fact, our estimation suggests that young people are paid roughly 23% less than their productivity. We thus argue that the youth wage subsidy proposal has no empirical basis in South Africa. The youth wage subsidy also has no empirical basis internationally, as demonstrated by the literature that National Treasury (and the Democratic Alliance) use to support the subsidy.

The seventh reason why we oppose the youth wage subsidy is that National Treasury (and the Democratic Alliance) incorrectly assumes that the wage is the major constraint to job-creation. The emphasis on the empirically unsubstantiated gap of an entry-level, or minimum wage, that is above productivity lies at the heart of National Treasury (and the Democratic Alliance's) standpoint.

In the first instance, the vast majority of young workers who fit the characteristics of many of the unemployed do not enjoy the statutory minimum wage. In other words, the minimum wage is not a binding constraint. Secondly there is no empirical basis to create a causal link between the extent of coverage by collective bargaining agreements and youth unemployment.Countries with high union and bargaining coverage do not necessarily exhibit high youth unemployment rates, the issue has more to do with economic structure and the role of the state in the economy.

The ninth reason why we oppose the youth wage subsidy is that it will simply increase the mark-up of firms without increasing employment. As we have argued above, the reasoning of National Treasury on the technical aspects of the youth wage subsidy is partial and incorrect. Given goods demand, it is clear that a wage cut for employers will simply raise the profit margin without increasing output. No firm will add labour simply because the wage has been cut for it, i.e. no firm will add labour beyond what is required to meet the demand for its goods at a given price. There is no firm that employs workers for charity or out of good heart. The law of the profit reigns supreme at all times, not social and political considerations.

The tenth reason why we oppose the youth wage subsidy is that it does not contribute in addressing the underlying causes of the youth unemployment problem. In fact the youth wage subsidy may exacerbate the triple crisis of poverty, unemployment and inequality. To think that our proposals are "long-term" in nature is to fail to understand their practical nature. The basic education system funnels 400 000 young people every year into the labour market. What is required is a national effort to drastically expand the education and training opportunities of these young people. The youth wage subsidy proposal is like taking out water from a highly leaky boat, using a small leaky bucket.

The COSATU position uses the National Skills Development Strategy III as a point of departure, by calling for expansion of the FET sector to accept 1 million learners per annum by 2014, compared to the current 400 000 per annum. This will in turn reduce the youth labour force, by extending their stay in the education and training system, so that they acquire basic and high-level cognitive skills (as the Germans and now the Australians are doing). Then state-owned enterprises, agencies and departments must stand ready to absorb these young people into practical training and provide work experience, especially given the massive infrastructure backlogs and maintenance that has to be done.

The private sector can do the same, without being given wage subsidies, but policies must be in place to support industrialisation and agriculture. The state must have capacity to plan and forecast for the numbers of young people who enter the post-school system and exit it, and ensure that no one falls through the cracks. This is what the German system does; it does not encourage young people to enter the labour market unskilled, it prepares and empowers young people to have solid career paths in the workplace. In South Africa, there exists a huge possibility that, even if the subsidy is re-designed to improve job-search, young people in desperation may choose to go straight into employment than increase their educational attainment. This has serious long-term impact on the career prospects for young people.

The South African Socio-Economic Context

Youth unemployment is obviously the dominant form of unemployment in South Africa, and has become a global phenomenon. The vast majority of the unemployed, an estimated 72%, are young people between 15-34 years of age; 43% of the unemployed are new entrants into the labour market and are therefore young, and 42% are between the ages of 25 and 341. Given the age profile of the unemployed, it is clear that youth unemployment is not a sectoral problem. Resolving youth unemployment is in fact almost equivalent to resolving unemployment in general. It therefore goes without saying that the line between youth unemployment and unemployment in general is extremely faint.

In terms of profile, 68% of the unemployed have been unemployed for more than a year, 60% of the unemployed have less than secondary education, 33% have completed secondary education but have no tertiary education, 95% of the unemployed do not have tertiary education. In addition, 60% of the unemployed have either never worked in their lives or have not worked in the past 5 years. This situation shows that South Africa may be in an "inequality trap". Unemployment tends to disproportionately affect Africans, women, people in rural areas, and those with less than Std 10.

In confronting this problem of youth unemployment, it is important to reflect on the current working conditions of young people. The ILO also reports that 30% of the employed young people in South Africa earn $2 a day, which is roughly R15 a day2. The UNDP's Human Development Report (2010) mentions that 44% of South African workers live on less than $1.25 a day, which is less than R10 a day. There is further evidence to show that the increased employment of young people over the past 10 years has been accompanied by increasing poverty among workers. The fall in poverty nationally has been due to social grants rather than the increase in employment that we have experienced3.

It is also important to reflect on the experiences of young workers whose characteristics are the same as those of the majority of the unemployed youth; they are African, are less than the age of 34, and have less than secondary education. Putting aside the other conditions of work, such as casualisation and labour brokering, their wage is less than R1 250.

A more detailed analysis of the average wage of young people with these characteristics shows that on average, these young people earn R940 a month4. This low wage level that is currently earned by young workers without secondary education shows that the problem of unemployment faced by these young people has nothing to do with wages, as the majority of them earn less than the minimum wage to begin with. In this regard the idea of a youth wage subsidy has no material basis in South Africa.

There is little training that is taking place in the sectors that have been creating employment in the past decade. Over the 7 years prior to the crisis, StatSA reports that approximately 1.9 million jobs were created in the South African economy. More than half of these jobs were lost as a result of the global economic crisis, especially in 2009. The sectors in which many of these jobs were created are the wholesale and retail trade sector, private services and construction. The New Growth Path states that the informal sector, agriculture and domestic work contribute a third of all employment, more than two thirds of working people earn less than R1 000 a month and one in five employed African women is a domestic worker5. Very little skills development and training is taking place in these sectors.

In fact the Commission on Employment Equity Report (2010) observes that for professionally qualified workers, skills development and training remains biased towards Whites, who command 61% of the skills development and training among those who are professionally qualified. Among skilled workers, the Commission on Employment Equity Report (2010) says that: "it is evident that private sector employers continue to invest more training on Whites than on other population groups.

If there was willingness on employers to empower Blacks, it would have been evident in the training provided. This therefore suggests that employers are not utilising their training strategically to ensure that they empower the under-represented groups to ensure their upward mobility within the workplace"6. Unfortunately, the Commission on Employment Equity Report does not provide information about skills development and training of the unskilled segment of the workforce.

National Treasury's (and the Democratic Alliance's) Main Arguments for the Youth Wage Subsidy in South Africa

On why young people are unemployed, the National Treasury provides the following reasons:

  • Employers look for skills and experience, they regard unskilled inexperienced jobseekers as a risky investment
  • Education is not a substitute for skills and experience. Schooling is not a reliable signal of capabilities, and low school quality feeds into poor workplace learning capacity.
  • Given the uncertainty about the potential of school leavers, employers consider entry-level wages to be too high relative to the risk of hiring inexperienced workers

The main motivation for the youth employment subsidy is "the gap between productivity and real wages for young workers", which is "an important constraint on job creation". A youth employment subsidy:

  • Reduces the financial costs or risks associated with not knowing the productivity of the person to be employed
  • Could help to make the training of young workers more affordable to employers, particularly smaller employers
  • May encourage more active job-search because youths believe that they are able to find work

National Treasury argues that the problem with South Africa is the "system of sectoral minimum wages [which] may have contributed to low levels of youth employment through pushing up the cost of entry-level workers"7. National Treasury further states that the average minimum wages in South Africa is about 62% the average formal sector wage, which is almost twice the international norm. This "exacerbates the implicit gap between entry-level wages and productivity, and hinders the hiring of younger workers".

There are side-effects that are associated with the youth wage subsidy. The National Treasury identifies a) deadweight loss, i.e. a situation where a subsidy is paid for unemployed persons who would have been hired without the subsidy. Note is taken that "deadweight loss is higher for general wage subsidies because targeted or marginal/incremental subsidies place limits on eligibility, b) substitution effects, where firms replace unsubsidised with subsidised workers, c) displacement effects, where firms with more subsidised workers crowd out those with less subsidised workers and d) destructive churning, which is a situation where employers fire young workers at the end of the subsidy period, or simply recycles subsidised workers and e) stigma effects, i.e. those that are from the subsidy system are stigmatised.

In relation to these side-effects, the Treasury document argues as follows: a) On substitution effects, the document argues that this "is unlikely to be substantial in the South African context. Young inexperienced individuals are not substitutes for experienced workers. There is little business sense behind replacing good experienced workers...with an inexperienced, young workers whose productivity is unknown, b) On replacement of existing workers, National Treasury argues that the youth wage subsidy is introduced "during the upswing (as evident in the current labour market data) and as such introducing the youth employment subsidy during the economic recovery is more likely to accelerate new hiring than result in replacement".In addition those already employed qualify for part of the subsidy and this should limit replacement, c) On the deadweight loss, Treasury argues that "the magnitude of the employment challenge in South Africa suggests that a temporary (!?) high deadweight costs may be an acceptable cost of stimulating youth employment, and d) On destructive churning, Treasury argues that "it is lousy business to fire good workers", by implication this is not likely to be a problem8.

In short, Treasury argues that the perceived gap between productivity and the entry-level wage is a major impediment to job-creation for young people. A subsidy to employers to close this gap will go a long way in reducing youth unemployment. The concerns over substitution, displacement, destructive churning and deadweight losses though important, are not significant. This is the gist of the argument in National Treasury's document on the youth wage subsidy.

The International Evidence on Youth Wage Subsidies

In this part of the paper, we show that the international literature that is marshalled by the National Treasury (and the Democratic Alliance), far from supporting their proposal on the youth wage subsidy, actually casts serious doubt on it. In some instances the literature is flatly opposed to a youth wage subsidy while in other cases the literature reports that wage subsidies have delivered mildly positive results under extremely limited and sometimes irrelevant conditions compared to our South African situation.

The United States: National Treasury relies on a paper by Katz (1996) to make a case based on the US. According to Katz (1996), the US has been experimenting with wage subsidies for more than 30 years. Wage subsidies came in the form of the New Jobs Tax Credit, the Targeted Jobs Tax Credit, and some aspects of Title II of the Job Training Partnership Act. To quote Katz in detail: "Two early attempts were the contract component of Job Opportunities in the Business Sector (JOBS), a late 1960s-early 1970s effort to train and hire disadvantaged workers; and the WIN tax credit (WINTC), a 1970s subsidy paid on the wages of AFDC recipients enrolled in the work-incentive program. The JOBS program started as a voluntary effort by the National Alliance of Business (NAB) to place young and less-educated workers from impoverished backgrounds in private employment"9.

Katz (1996) notes that OECD-style targeted subsidies "can generate incentives for higher turnover and disproportionately benefit sectors with high turnover and more variable employment", while subsidies that are given to struggling firms "can create moral hazard with firms; threatening to lay off to gain a subsidy". In relation to the United States experience, Katz presents disappointing evidence: "Unfortunately, the Joint Training Partnership Act evaluation indicates this same strategy has no impacts on the earnings or employment prospects of out-of-school disadvantaged youth. A mixed strategy providing job search assistance, job development, and a temporary wage subsidy seems to be a successful approach to substantially improving the earnings of disadvantagedadults (particularly adult women). But, like other non-intensive strategies that have been evaluated, it does not appear very effective for out-of-school youth from poor families". We note that the paper by Katz (1996) is listed in the references of the National Treasury document and is used in p.29 of the National Treasury document as part of international evidence that youth wage subsidies work in the US to help "economically disadvantaged youths"!

The United Kingdom: The Department for Business, Innovation and Skills issued a very strong communiqué against youth wage subsidies. It states, "The UK Government has however examined the economic case for general UK wage support for all employers in-depth. Our conclusion is that this is not a feasible, cost effective or sustainable option for us. We also know, based on our experiences of the 1970s, that these schemes can create distortions and prevent companies taking necessary action to restructure or retrain to ensure their survival and success in the future, leading to failure and redundancies when the subsidy is removed. They can also provide perverse incentives for other sectors and companies to ‘bargain' for subsidies"10. The UK government further says: "Work by the OECD concludes that wage subsidy schemes are extremely costly. This is because of high deadweight costs as many eligible firms would in fact have retained workers without the subsidy".

Van Reenen (2003) is the study that National Treasury relies on to make the case for the UK11. However, it is clear from the reading that the "New Deal" is a package of interventions in which the wage subsidy is just one component. In fact National Treasury itself reports that the wage subsidy component accounted for only 13% of the take-up by young people in 1999. The later study by Van Reenen (2003) reports the take-up to be just 20%. Van Reenen (2003) mentions that the "New Deal" increases the probability of employment of young people by 20%, much of this is due to the wage subsidy, but he also notes that "at least a fifth of the effect is due to enhanced job search".

In fact, in the text van Reenen (2003) does not mention how much the "pure subsidy effect" is, instead he says: "The "true" effect of job assistance is likely to be higher as some of those obtaining subsidised jobs would have obtained them even in the absence of a subsidy, despite the best efforts of the employment service to minimise this deadweight". Furthermore the effect of the subsidy crucially depends on the productivity gains from training and work experience, an issue that has not been investigated yet. Given the uncertainty surrounding its effects, it is no wonder that the UK government is sternly opposed to the wage subsidy and, as van Reenen (2003) says, this part of the "New Deal" faces heavy cuts.

Argentina: the National Treasury document relies on the paper by Galasso, Ravallion and Salvia (2001)12. The Argentinean experience was basically an experiment that was conducted in two adjacent Argentinean towns which suffered massive retrenchments, but in which there was temporary work already in place in the form of an infrastructure development programme. This was therefore not a national programme (as the National Treasury document implicitly wants us to believe). One thing about the subsidy programme is that it is voucher-based, the targeted person is given a voucher, uses it to look for employment, the employer then pays the voucher-holder a discounted wage and the discount is paid directly by government to the voucher holder. Clearly this programme has elements of both supply-side, i.e. the unemployed are encouraged to search for work and the demand-side, i.e. the discounted wage cheapens labour for employers. However in order to claim the subsidy firms have to formally register the workers.

The authors find that the "take up of the wage subsidy by hiring firms was low" because of the costs associated with formally registering employed workers. In Argentina once a worker is registered, firms have to pay social security and cannot fire workers once the subsidy is over because there would be severance penalties13. In the light of this, the authors state: "So the impact of the voucher was clearly not through access to the wage subsidy by firms...The voucher may well have encouraged workers to make more effort to find work.

By this interpretation, the voucher had an "empowerment" effect, in making these workers -notably young and female workers - more confident in approaching employers". Although they acknowledge the cost effectiveness of the vouchers, the authors state that "clearly, Proempleo did not succeed in achieving a major transition to private sector jobs amongst workfare recipients in the study area". Whatever gains in employment that they report, the authors are clear that the "empowerment" effect was dominant, and they raise complications if the experiment is to be scaled up to national level. These conclusions hardly support National Treasury's and the Democratic Alliance's position, less so their proposed subsidy!

Australia: The National Treasury relies on the paper by Richardson (1998)14. This paper discusses the Special Youth Employment Training Program "which was introduced in 1976...until it was superseded in December 1985 by Jobstart, a wage subsidy for both youth and adults, which has continued until the present". Richardson mentions that the Program "required both referral by the Employment Service and acceptance by the employer", a feature which is absent in National Treasury's youth wage subsidy. This "referral" implies that there is some selection bias in the Program. In order to evaluate the effect of the subsidy, one needs to split the unemployed into two groups, those who are part of the subsidy and those who are not. Then one has to ensure that the characteristics of those that are selected into the subsidy program are the same as those that are not part, otherwise the effect of the subsidy on the participants will be confounded by differences in characteristics. This is found by Knight (2002) to be a problem with Richardson's (1998) study.

There are differences in post-school qualifications, work history and unemployment duration among those who were on the subsidy and those who not15. Taking all these and other problems into account, Knight (2002) concludes that "In all variations, the research here found positive impacts of SYETP, even if some were not statistically significant". The paper by Richardson (1998) therefore over-estimates the impact of the subsidy. Recent initiatives to introduce the wage subsidy in Australia reveal that there have been major problems in the past, there is now emphasis on preventing firms from retrenching young people after the subsidy has expired, heavy inspection to ensure that training takes place, that no worker substitution occurs, etc. But all these measures are likely to fail as they are administratively cumbersome.

Interestingly, the more recent efforts to combat youth unemployment in Australia do not mention youth wage subsidies and there is a discernible shift towards institutionalised education in order to lay a foundation for the acquisition of basic cognitive skills by young people. For example, the OECD (2009) notes that "Australia is indeed characterised by a relatively low retention rate in education beyond age 16, as compared with many other OECD countries. Hence, too many youths still lack the basic skills needed to embark on a successful long-term career in the labour market". Instead of thrusting young, ill-prepared school-leavers into the labour market, the main thrust of the strategy is to "raise the average level of educational attainment...develop vocational education and training within the school system, as well as improve the school to work transition".

This point was echoed by Dorrance and Hughes (1996):"The high level of basic cognitive skills that have to be acquired during school and post-secondary education to make further training possible, however, introduces a new dimension. Girls and boys who do not acquire appropriately high educational levels, will be cut off from mainstream career development...Australians are becoming increasingly aware of the need to invest in their own education"16. Thus, the weaknesses of the school and post-school education system in South Africa will impede workplace training and limit career development of those whose wages are subsidised, especially considering their educational attainment.

Colombia: The National Treasury here relies on a publication by Colombia's Ministry of Social Protection. We could not get hold of this paper. However a reading of the state of affairs in Colombia seems to suggest that youth unemployment remains a serious problem. In addition, there seems to have been massive problems associated with lack of career development of young people in the Colombian labour market. This problem is so serious that commentators in Colombia are concerned that youth may increasingly get disillusioned by "a system of broken promises".

It is difficult to get literature that analyses the effect of subsidised employment for Colombia. However there seems to be a problem of low education attainment in Colombia, affecting the effectiveness of training. Schaffner (2001) notes that "the cost of imparting skills to workers through job training is thought to be higher for workers with less, or lower quality, formal education"17. This applies to more than 60% of the unemployed youth in South Africa.

A paper by Attanasio et.al. (2009) conducts an evaluation of a subsidised vocational training programme in Colombia called the "Youth in Action" program. The authors find that: "Training offers increase the probability of having a formal sector job by 0.053. The results also show that the offer of training increases the probability of having a written contract by 0.066"18.

In other words, those who are not part of the program have a 6.6 percent less chance of getting written employment contract. These findings are to us very marginal or small, especially if one were to consider the economy-wide effects of such a programme if it were to be rolled nationally. It should also be noted that this "Youth in Action" program "consisted of 3 months of classroom training and 3 months of on-the-job training". This component is extremely important because it shows that the Colombian program is vastly different from the "cold turkey" approach of National Treasury's and the Democratic Alliance's youth wage subsidy19.

Despite this advanced design, the Colombian program appears to just deliver a marginal effect. The other problem is that in Colombia workers tend to have a high frequency of changing jobs, which is thought to be due to the prevalence of short-term job opportunities compared to long-term jobs, a matter of economic structure20. Furthermore, despite valiant efforts with subsidies, Colombia seems to be failing to career-path young people. This is evident in the concerns relating to the new programme for young people: "A youth labour law should tackle cultural and deep-rooted practices that work against any intention of guaranteeing young people a better future through decent work. As a number of companies experience first-hand, a lot of the youth, especially those coming from poverty, simply look for a job, but do not really build a career"21.

Turkey: The National Treasury document relies on the paper by Betcherman and Daysal (2009) to make a case for Turkey. We use the version that was published in 201022. According to the authors, "Law 5084 (2004), which covered 37 provinces and those given by Law 5350 (2005), which expanded coverage to 13 additional provinces, and modified qualification rules and subsidy amounts relative to the first scheme. Both laws subsidized employers` social security contributions, employee personal income taxes, energy consumption and land".

The aim of government was to encouraged investment and employment in low income provinces in Turkey. These subsidies are not youth wage subsidies, but general employment subsidies through which firms can claimdiscounts on electricity consumption, social security contributions, land taxes, etc. They are thus not comparable to the type of subsidy that the National Treasury and the Democratic Alliance want to introduce. For example, the authors use a method that does not focus on individuals, i.e. their characteristics, but focuses on differences and treatment of provinces that were targeted versus those that were not targeted.

The results for Turkey are also not encouraging. The authors say that "the evidence suggests that the dominant effect of subsidies was to increase social security registration of firms and workersrather than boosting total employment and economic activity".Furthermore they find that a) "the gains in employment and number of firms correspond to a surge in formalisation, rather than to real gains in economic activity...", b) "our results also suggest that implementing the subsidies can be a costly way to increase employment" and c) the deadweight costs of the programme were so high that "the government ended up paying for nearly the full cost of employment created". Thus, in relation to Turkey, because firms have to be formally registered to claim the subsidy, including registering the newly employed, what the subsidies have done is to increase formalisation, but at an extremely huge cost! Thus, once again the very study that Treasury (and the Democratic Alliance) use to support the youth wage subsidy suggests that this strategy is not the best strategy to boost employment.

Estonia and the Czech Republic: The paper by Wilson and Fretwell (1999) that is mentioned in the Treasury document is not listed in the references, and so we could not find it. The one we could find deals with Public Service Employment and not subsidised employment23. We then looked at the paper by Fretwell et.al (1999), who find that the impact of the training programs "had a small but positive impact on employment, except for Turkey where the findings are not significant or were negative"24. Furthermore these authors find that the impact "tended to dissipate over time for the short Czech training programs". Overall the impact of the subsidy in the Czech Republic is not supportive of the Treasury (and the Democratic Alliance) position.

In relation to Estonia, Leetma and Vork (2003) say that "the level of the wage subsidy is 100% of the minimum wage during the first 6 months and 50% of the minimum wage during the next 6 months of her/his employment period"25. In comparing those who participated and those who did not, the authors admit that "it is not clear, whether those participating in ALMPs have better labour market prospects on average in our sample". They nevertheless run regressions, and conclude that "The study shows positive and statistically significant impact of the active labour market programmes on employment probability. The effect turns to be positive 4 months after the end of programmes and it reaches 7-8% during the year after the program". The deadweight costs of the programme are not evaluated and neither are the possible substitution and displacement effects. However, still, where the probability of employment is improved, it is quite small to justify the rolling out of such programmes on a large scale, especially given that for 6 months government paid the full minimum wage.

Denmark: The National Treasury here relies on the paper by Rotgers and Arendt (2010)26. The authors report that in Denmark "the subsidy covers approximately 50% of the wage and is constant during the subsidised period. The total duration of subsidised spells cannot be higher than one year". There are conditionalities on how firms can employ and benefit from subsidies. For example, "firms with 1-5 total employees might employ at most 1 subsidised employee, firms with 6-50 employees might hold at most 1 subsidised employee per 5 ordinary employees, while firms with more than 50 employees, can afford at most 1 subsidised employee per 10 ordinary employees...Finally, in order to hire a subsidised employee the firm requires the approval of the employees' representative". Two points have to be made. Firstly, the Rotgers-Arendt paper is not about youth employment subsidies, it is about general employment subsidies. Secondly the paper does not focus on individual characteristics of the unemployed, rather it focuses on the characteristics of firms; it investigates whether wage subsidies have an impact on ordinary employment of subsidised firms. This paper can thus not be used to make a case for youth wage subsidies as such, because it does not address itself to the characteristics of the unemployed, their age profile, work history, educational background, gender, race, location (urban or rural), duration of unemployment, etc. These factors are extremely relevant in our situation.

Rotgers and Arendt find that "hiring a subsidised employee has a significant positive average employment effect on the subsidised firm already 1 month after the beginning of the subsidised contract. As time passes, the positive effect on the firms' ordinary employment increases suggesting that on average subsidised employers tend to hire the subsidised employee on ordinary terms or use subsidy to financing the hiring of other individuals on ordinary conditions". Much as the paper is not quite relevant to our situation, it is however not clear how Rotgers and Arendt address the identification problem. Given that firms have to employ subsidised workers in a determinate ratio to ordinary workers, it is not clear whether firms "add on" subsidised workers in the course of raising ordinary employment. For if firms "add on" subsidised workers, then it is difficult to draw a causal link between the subsidy and ordinary employment. Thus, it is difficult to tell if subsidised employment increases as a result of an increase in ordinary employment, or whether ordinary employment increases as a result of subsidised employment. In our view, the former is likely to be the case, since the latter may exert exorbitant costs to firms. It would not make sense to add, say, 5 more ordinary workers in order to access a subsidy for one additional worker.

Germany: In relation to Germany, the National Treasury relies on a paper by Jaenichen and Stephan (2007). According to the authors "The EGZ for hard-to-place persons could regularly account for as much as 50 percent of the monthly wage or salary and continue for at most 12 months. These limits could be exceeded in exceptional cases. As a special feature of the EGZ, a follow-up period of further employment is obligatory after the expiration of the subsidy.

If a person hired with an EGZ is dismissed within this period for reasons attributable to the employer, the employer can be asked to reimburse part of the subsidy". The paper does not detail the socio-demographic characteristics of the "hard-to-place" workers, which makes it difficult to compare with a youth wage subsidy programme.

However, looking at the figures provided by Jaenichen and Stephan, it is clear that the wage subsidy accounted for just 3% of all the transitions from unemployment. As the authors note: "looking at aggregate figures, exits into a wage subsidy programme are of limited importance for leaving unemployment in Germany. Furthermore, compared to nonsubsidised jobs, subsidised jobs only account for a small fraction of transitions to employment".

The German EGZ program is not a youth wage subsidy programme. As Boockmann et.al. (2007) indicate: "In 1998, the German government introduced the Integration Supplement for Older Workers(Eingliederungszuschuss für ältere Arbeitnehmer, EGZ)hiring subsidies targeted at workers aged 50 and above"27. If this is the case, it completely renders the EGZ programme irrelevant to what the National Treasury (and the Democratic Alliance) propose. Jaenichen and Stephan state "Our results show that wage subsidies may increase the employment prospects of supported workers to a considerable amount. For previously unemployed individuals, three years after the start of the programme, the share in regular employment is from 25 to 42 percent higher in the treatment group than in the matched control group. A comparison between groups of unemployed persons taking subsidised employment with matched control groups of individuals moving directly into unsubsidised employment indicates that differences in the employment prospects are rather small after three years". So, the benefits of the subsidy disappear after three years.

But Boockmann et.al. (2007) find the opposing result: "Estimates from duration analysis of the transition rate show that, for the population as a whole, the probability of leaving unemployment and entering employment is unaffected by the availability of hiring subsidies...Further analysis shows that the ineffectiveness for most population groups is due to deadweight effects: an increase in subsidized hirings is accompanied by a decline in unsubsidized new employment". Similarly Schünemann et.al. (2011) find that in Germany, "The results suggest no significant effect of the subsidy on exit rates out of unemployment or employment stability. Employment rates up to three years after eligibility show no significant improvement. In conclusion, our findings are in contrast to previous empirical results justifying such policies". We need not add any further comment in relation to Germany's experience, upon which the National Treasury and the Democratic Alliance base their proposals.

Poland: The National Treasury refers here to a 269-page report by O'Leary (1998)28. In that report, O'Leary evaluates the entire spectrum of labour market policies, one of which is a retraining programme. Treasury's document seems to focus on the Intervention Works Programme (See p.30 Box 3) to derive its support. But we focus on two aspects of the active labour market policies for a broader perspective on Poland: retraining and intervention works.

O'Leary states that "in retraining, unemployed workers are given additional short-term job skill training to make them ready to fill job openings in the region. Retraining participants receive a stipend which has a 15 percent premium over the (UC) benefit. In relation to intervention works, O'Leary states that these are "much like public works except that projects may not compete with private companies and the wage paid by grants can be no more than the unemployment compensation benefit.

Projects may be operated by either public agencies or private companies. There may be no intervention works contracts given to employers who have laid off significant numbers of workers in recent months. There are also incentives for employers to permanently retain workers. After the end of an intervention works project, which may last up to 6 months, employers can receive wage subsidies for retained workers amounting to up to 150 percent of the national average wage. Intervention works operates essentially as a wage subsidy program".

O'Leary (1998) finds that "Retraining resulted in more people (12 percentage points) getting into regular nonsubsidized employment and a 23 Zl. gain in average monthly earnings. Retraining was more effective for prime-age workers, with a non-vocational background, who had occupations which could not be easily categorized into broad occupational groups, were not previously long-term unemployed, had either very short or rather long prior employment history, and lived in voivods with a high unemployment rate. It was also found that short-term skill focused retraining was most effective, and there was some evidence that retraining provided by private firms was more effective. It is better if retraining is provided by an adult education or other firm engaged in normal industrial activity rather than having training provided by an employment organization or having another labour-related group serve as the trainer". In other words, private service providers who provide training are bad for the labour market in Poland.

The impact of retraining was lowest, almost 5%, among blue-collar workers. Thus, retraining in Poland was more effective for the short-term unemployed and people with some employment history. These characteristics rule out 68% of the unemployed in South Africa, who suffer from long-term unemployment. This means that the Polish experience is not that useful. Even an analysis of the performance of other active labour market programs in Poland shows that they are better than re-training. In terms of intervention works O'Leary reports massive success, with re-employment of those who participated ranging from 52% to 68%. O'Leary reports that "among program participants a whopping 59.7 percent were retained as regular employees by the intervention works program operator after government funding of the project ended"29. Clearly, if things ended here this aspect of Poland's active labour market policy would be worth studying in more detail.

However, O'Leary (1998) provides some further, devastating information: "A large and statistically significant difference appeared indicating that those who are not long term unemployed benefit appreciably more from intervention works in terms of re-employment than those who were long-term unemployed. Indeed, those who were not long-term unemployed appeared to capture all the gains in terms of employment effects, while those with more than 12 months of registered unemployment generally had their reemployment success hurt by participation in intervention works. While not statistically significant, the negative impact on earnings was also greater for the long term unemployed". Thus the intervention works type of programme would "hurt" the re-employment chances of the 68% of the South African unemployed. In short, given the long-term structural nature of unemployment in South Africa, the Polish experience is not useful and in fact, far from supporting the National Treasury and the Democratic Alliance's view, the Polish experience suggests that wage subsidies should not be adopted in South Africa.

Sweden and Slovakia: Perhaps we do not have to spend a lot of time on Sweden. Larsson (2000)30 and Larsson (2003)31 are the studies to which National Treasury refers. But as Treasury itself reports, the Swedish experience does not favour youth wage subsidies. Sweden experienced negative effects on employment and earnings of participants, almost zero long run effects on employment after the subsidy has ended and negative employment effects after one year of participating in the subsidy programme. In relation to Slovakia, National Treasury relies on a paper by van Ours (2000). We would not go into detail here since the Treasury itself also acknowledges that the Socially Purposeful Jobs, which provides a "wage subsidy for the employment of the target group", had a "negative effect on the job-finding rate". These observations from Treasury's own literature hardly support its (and the Democratic Alliance's) proposals for a youth wage subsidy.

We think that South Africa can learn a lot from a careful study of the German labour market policies. The emphasis in Germany has always been on ever raising the basic cognitive skills of young people, providing them with a range of options and career paths which are embedded in a well-structured institutional context in which the state plays a central role. Australia seems to have awoken to the reality that without developing basic cognitive skills in an institutionalised context, on-the-job training is likely to be extremely futile and costly32.

Gross (1997), in comparing the Canadian and German systems, reaches the following conclusion: "To summarize, the German education system is very structured and less gifted students have little access to purely academic options. However, they can pursue advanced training in the apprenticeship programme and in vocational schools. As a result, the option least favoured by teenagers is to enter the labour market as an unskilled worker"33. This preference by teenagers appears to be a global phenomenon, as we have seen in the UK in relation to the "New Deal" and in Australia, and it makes sense; nobody wants to enter into dead-end job. The youth wage subsidy proposal by National Treasury is premised on the failed Canadian-style private-sector-centred labour market intervention, instead of the more successful German-style state-centred intervention, which emphasises vocational, technical and academic training linked supported by an active state-directed workplace experience component.

The analysis of the "New Deal" in the UK shows that young people prefer skills development and education over entering the labour market without institutionalised, formal training: "Out of the four options, education and training has been the most popular (about 40% of all those who had joined the New Deal options by the end of April 1999 chose education and training). The employer's option [i.e. subsidised employment] had a much lower take-up than anticipated (only 20% of all those in options)"34.

Such is the state of affairs in relation to international evidence. The conclusion that can be drawn from the above analysis shows that the proposals from National Treasury and the Democratic Alliance have extremely narrow and limited, if no, empirical basis in international evidence. The very same literature that they use to justify their proposal does not support the use of youth wage subsidies. This means that they have not carefully read the literature, and neither have they bothered to analyse the strict, extremely limited and more often than not irrelevant, conditions under which wage subsidies appeared to have had extremely small positive results, at huge cost to the public purse.

The Economics of the Youth Wage Subsidy Reconsidered

In p.29 the National Treasury provides a "technical explanation for why an employment subsidy increases job creation", in Box 2. The explanation is a simple neo-classical one. Labour demand is downward-sloping and labour-supply may be flat or upward sloping. In the case where there is mass unemployment, the labour supply is horizontal. Indeed, it can be shown from profit-maximisation assumptions that a wage subsidy shifts the demand for labour to the right; for the same wage employers would employ more workers. But there are at least two complications here, with the Treasury's neo-classical diagram.

The first complication is that the wage that appears on the vertical axis is both mapped on to the demand-side (employers) and supply-side (workers). Workers receive the same wage with or without the subsidy, but employers pay two sets of wage rates: one for the unsubsidised and the other for the subsidised. It is therefore invalid to map labour demand (and supply) as if these two quantities of labour are of the same quality. To be specific, what the wage subsidy seeks to achieve is to stimulate more employment for the targeted group of workers because of certain characteristics that make them normally unemployable.

Therefore, the supply and demand that should be in Box 2 of the Treasury document should be the supply and demand for the "targeted labour". Now, since the "targeted labour" is massively unemployed, this labour has no power to set the wage, i.e. the"targeted unemployed" are price-takers in the labour market. This means that whatever wage falls out of the profit-maximisation calculus of the firm, these targeted workers will have to accept it.

Accepting the "targeted labour"demand by firms to be downward-sloping in the "targeted wage", it follows that the effect of the wage subsidy is to decrease the wage paid by firms for targeted workers. Consequently the firm will move along its "targeted labour" demand curve. In the National Treasury document the firm experiences a shift in its demand curve.

This shift is untenable, since it assumes that the subsidy's discount factor on the wage "boosts" the marginal product of labour when in fact, such a subsidy discount decreases the wage paid by the firm. This point is illustrated by Figure 1. In Panel A we reproduce Treasury's incorrect diagram, where the subsidy discount (1-s) is clearly indicated but is nevertheless incorrectly thought to shift the demand for labour to the right, so that with a subsidy, firms demand.

Figure 1: An Illustration of the Effect of the Wage Subsidy or Why Treasury's Diagram Is Wrong

A The Incorrect Diagram for Targeted Subsidy       B. The Correct Diagram for Targeted Subsidy

5.4 On the other hand, Panel B assumes that firms hire the "targeted labour" at some entry level wage which is thought not to reflect marginal productivity, e.g. the minimum wage that the Treasury document alleges distorts the labour market. As the National Treasury document consistently maintains without any empirical substantiation, the "large gap between real wages and productivity for young people" has to be closed by the subsidy. Thus, the subsidy reduces the wage rate from the minimum wage to the subsidised entry-level wage, so that, where is the subsidy rate.

In our view, the quantity demanded for the targeted labour increases to at Point B on Panel B. However, we also note that in a dynamic context, where economic growth exceeds the growth rate of targeted labour's employment, the slope of the labour demand curve will become steep, swivelling targeted labour demand from the solid line to the dotted line. Under these realistic circumstances, wages subsidy is likely to have only marginal effect, by just increasing the targeted quantity of labour demanded.

5.5 The second complication with this neo-classical analysis is that for more labour to be demanded there should be an increase in output that must be accommodated. This flows directly from the notion of the production function which states that the more labour is employed, the more output will be produced, all else being the same. But for this to be realised the output supply curve must shift, thereby delivering a fall in output price.

Clearly this process is in reality complicated, especially when the economy is dominated by oligopolies, monopolies and is imperfectly competitive. Rather, the reverse is likely to be the case, i.e. given the demand for goods and services, a cut in the wage cost through the subsidy will increase the profit margin by shifting all cost curves down, without having any impact on output supply. In other words, the wage subsidy will simply increase the price mark-up over labour cost by the amount of the subsidy rate. Therefore the subsidy is not going to deliver the number of jobs that are projected by National Treasury (and the Democratic Alliance).

Why Then COSATU Opposes the Youth Wage Subsidy?

We have shown that the evidence that is contained in the National Treasury document on youth wage subsidies, or even employment subsidies, leans towards the COSATU position. Based on Treasury's own literature which they used to propose this ill-conceived policy, we have shown that there are no empirical grounds for this policy. The irrelevance and the likely wasteful effects of this proposal have been amply demonstrated in many cases.

International literature shows that the wage subsidy idea is extremely costly and wasteful, with massive deadweight losses. The overwhelming literature does not venture to estimate the economy-wide impacts of such initiatives, since in most cases they are implemented in localised places and not at national level. Those that have a more or less national scale, tend to be ineffective. Examples here are the case of Turkey, where government ended up paying almost the entire wage.

The ILO (2011) reports that "research in various countries has shown that wage subsidies lead to combined deadweight and substitution effects of the order of 70-90% of the number of jobs created"35. The estimate by National Treasury puts the deadweight loss alone to be 58%, i.e. 58% of the jobs that would have been created without the subsidy will be subsidised (that is, if we believe National Treasury's estimates!).

The youth wage subsidy will have significant substitution effects. Firms will have an incentive to let go of existing workers in order to employ subsidised ones. The National Treasury document dismisses this concern on two grounds: a) "regulations around the dismissal of existing workers establish a legal framework the prevents this kind of substitution from occurring" and b) "there is little business sense behind replacing good experienced workers who have demonstrated their productivity and value to a firm with an inexperienced, young worker...".

In relation to a), National Treasury pretends as if it does not operate in South Africa, where the elementary rights of workers are violated on a daily basis. Government dismally fails to protect farm workers who are trapped in near slavery and neo-feudal relationships with their employers - also domestic workers, taxi drivers, despite the 2009 Manifesto commitment to improve enforceability of labour law to better protect farm workers and other vulnerable workers.

For example, the vast majority of workers do not enjoy the minimum wage regulation. But also more pertinent here, is the fact that only 29% of the workforce is unionised in South Africa, which opens up the rest, 71%, to abuse. In relation to b), we have demonstrated that unskilled workers are not being trained and are employed in sectors that require minimal skills. In addition, the existence of labour brokers who screen workers for employers also makes it easy to fire existing workers and get good ones on a subsidised basis.

The substitution effects are likely to be widespread, especially with the existence of labour brokers. This substitution will hit the unskilled and semi-skilled parts of the workforce the most. By this criterion, and using the 2011 industry structure tables, we find that at least 3.7 million workers are vulnerable to substitution in the South African economy. The tables also include a category of mid-level skilled workers, whose skills composition is difficult to ascertain. However, we can interpret the figure of 3.7 million as a minimum number of workers who are vulnerable to substitution.

Table 1: Number of Workers Who Are Vulnerable to Substitution in the Formal Sector (Unskilled and Semi-Skilled)

 

Absolute Numbers of Workers

Percent of the Workforce

Manufacturing

622 089

46.44

Mining

452 553

76.1

Construction

313 572

45.36

Wholesale & Retail

273 301

11.37

Transport & Storage

97 426

17.57

Finance

265 814

13.54

Community, Social and Personal Services

1 397 306

37.56

General Government Services

276 107

14.49

Total

3 698 168

 

The National Treasury document argues that the subsidy is introduced at the beginning of a recovery, so that substitution effects will be limited. But this argument is clearly not structural as it does not consider the skills composition of the sectors involved. Secondly, even if we concede that economic growth will limit substitution, what will happen is that deadweight costs would increase, because with economic growth firms would hire more young workers even if the subsidy is not offered to them.

Thus the economics is that a low rate of economic growth would induce high levels of substitution and low levels of deadweight, whilst a high rate of economic growth delivers low substitution and high deadweight costs. But in our case, given the structure of the economy, its skills composition and the composition of skills among the unemployed, it follows that deadweight costs and substitution effects are likely to be high with economic growth.

The third reason why we oppose the youth wage subsidy is that it does not guarantee that training and skills development will take place in the workplace, less so in the sectors where job-creation is likely to be created: wholesale and retail trade, personal services and construction. As we have noted from the Commission on Equity and Empowerment, little training is dedicated to black people, less so to the skilled segment of the workforce. The situation is worse for the unskilled, who are likely to be outsourced, casualised and employed through labour brokers.

What is even worse, the National Treasury does not want to mandate training: "the design of any potential employment subsidy may not want to mandate training alongside the subsidy since additional administrative burdens on employers may discourage take-up of the subsidy". This is indeed a problem, as we have seen in many cases, such as Argentina and Turkey. But the literature also says that these subsidies depend on training for success. Without mandating training, which is currently very minimal, it is clear that National Treasury relies purely on the philanthropy of the private sector, an expectation which runs against the daily experience of the vast majority in the South African workforce.

The fourth reason why we oppose the youth wage subsidy is that it will lead to the recycling of young people without training. In the literature they say young people will be fired once the subsidy ends. National Treasury (and the Democratic Alliance) dismiss this on the grounds that "it's lousy business to fire good workers". But the fact that businesses have moved drastically towards outsourcing, labour broking and casualisation shows that it is good business to have a workforce that is vulnerable and flexible.

The goodness of a worker matters to business as long as that worker can be employed in as low a wage as possible, to produce as much profit as possible. The goodness of the worker is thus subordinated to the power profit. Indeed there are many good workers that have been retrenched only to be hired under labour brokers, or as casuals. Because National Treasury does not mandate that business be held accountable for "recycling", and seeks to ensure that, during the subsidy period, these young people do not have recourse to labour protection, the proposed youth wage subsidy will produce massive "destructive churning".

A young worker under the subsidy would have hopped jobs many times when the subsidy ends, because the state is taking a hands-off approach in the process. Indeed the power of employers over these young people will be absolute. This has been noted to be a major problem in Australia and Colombia, hence the move away from subsidies towards formal training along German lines in Australia.

The fifth reason why we oppose the youth wage subsidy is that with major substitution and increased vulnerability of the workforce, there will be downward pressure on wages. Inequality will worsen as low wage workers replace those that have managed to capture non-wage benefits in their compensation. It can be shown that the increase in the mark-up due to the subsidy will raise the profit-share at the expense of the labour share. This therefore will not take us forward with the triple challenges. Indeed jobs have been created, but at the level of poverty wages. In additionpoverty is likely to rise, because employed workers with relatively higher wages will be replaced by many vulnerable low wage workers.

The sixth reason why we oppose the subsidy is that there is an underlying assumption that there is a gap between entry-level wages and productivity among young workers. It argues that youth wage rates are too high. However the National treasury document fails to compute this gap between the wage and productivity. With an average wage of R940 for those that fit the characteristics of at least 60% of the unemployed, it would be interesting to know what is expensive about this average monthly wage. We thus argue that the youth wage subsidy proposal has no empirical basis in South Africa.

The youth wage subsidy also has no empirical basis internationally. We have demonstrated that even the literature that National Treasury (and the Democratic Alliance) use is not in support of the subsidy. Where these subsidies seem to have worked is under conditions that of extremely limited relevance to our situation and the effect has been very marginal. Therefore, if international evidence is to be marshalled, as National treasury has done, it points towards the correctness of the COSATU position on this issue.

The seventh reason why we oppose the youth wage subsidy is that National Treasury (and the Democratic Alliance) incorrectly assumes that the wage is the major constraint to job-creation. The emphasis on the empirically unsubstantiated gap of an entry-level, or minimum wage, that is above productivity lies at the heart of National Treasury (and the Democratic Alliance's) standpoint. In the first instance, the vast majority of young workers who fit the characteristics of many of the unemployed do not enjoy the statutory minimum wage. In other words, the minimum wage is not a binding constraint. Secondly the extent of coverage of collective bargaining agreements is not related to youth unemployment. For example, as noted by Gross (1997), "union coverage in Germany is 90%; in Canada, it is below 40%. Bargained wages in Germany serve as minimum wages and therefore apply to all skill categories. There is a reward for experience, as special lower wage rates exist for young people (sometimes up to the age of 21)".

Yet, Germany has a lower youth unemployment rate than Canada. Indeed there is no empirical basis to link collective bargaining coverage, or unionisation, to the scale of youth unemployment. Youth unemployment has to do with the role of the state and the structure of the economy.

The National Treasury document attempts to attack collective bargaining so that it does not extend to young people, because it argues that collective bargaining pushes up youth wages above their productivity. Yet, advocates of the youth wage subsidy sternly deny that they seek to create a two-tier labour market system.

The ninth reason why we oppose the youth wage subsidy is that it will simply increase the mark-up of firms without increasing employment. As we have argued above, the reasoning of National Treasury on the technical aspects of the youth wage subsidy is partial and incorrect. Given goods demand, it is clear that a wage cut for employers will simply raise the profit margin without increasing output. No firm will add labour simply because the wage has been cut for it, i.e. no firm will add labour beyond what is required to meet the demand for its goods at a given price.

The tenth reason why we oppose the youth wage subsidy is that it does not contribute in addressing the underlying causes of the youth unemployment problem. In fact the youth wage subsidy may exacerbate the triple crisis of poverty, unemployment and inequality. To think that our proposals are "long-term" in nature is to fail to understand their practical nature. The basic education system funnels 400 000 young people every year into the labour market. What is required is a national effort to drastically expand the education and training opportunities of these young people. The youth wage subsidy proposal is like taking out water from a highly leaky boat, using a small leaky bucket.

The COSATU proposal uses the National Skills Development Strategy III as a point of departure, by calling for expansion of the FET sector to accept 1 million learners per annum by 2014, compared to the current 400 000 per annum. This will in turn reduce the youth labour force, by extending their stay in the education and training system, so that they acquire basic and high-level cognitive skills (as the Germans and now the Australians are doing). Then state-owned enterprises, agencies and departments must stand ready to absorb these young people into practical training and provide work experience, especially given the massive infrastructure backlogs and maintenance that has to be done.

The private sector can do the same, without being given wage subsidies, but policies to support industrialisation and agriculture must be vigorously pursued. The state must have capacity to plan and forecast for the numbers of young people who enter the post-school system and exit it, and ensure that no one falls through the cracks. This is what the German system does; it does not encourage young people to enter the labour market unskilled, it prepares and empowers young people to have solid career paths in the workplace.

The economics of this proposal is very simple. Youth unemployment is massive, so that in the extreme case young people take the wage that is provided in the labour market (mainly by employers). We assume that this wage is , which corresponds to the horizontal line labelled "entry-level wage rate, mass unemployment". It can also be shown that as the wage rate increases, the unemployment rate will tend to rise, given the level of aggregate demand. We thus have an intersection with mass unemployment at point A, where the youth unemployment rate is high at . Alternatively, young people may have some bargaining power, in which case the "youth wage bargaining curve" tells us that an increase in the youth unemployment rate will tend to weaken the bargaining power of young people, whilst a decrease in youth unemployment will tend to increase their bargaining power.

Figure 2: A Technical Explanation of the COSATU Proposal

COSATU's proposal is that the state must decrease the youth labour force by trapping young people longer into the education and training system, e.g. by increasing the additional registrations into FET's by 600 000. The impact of this will be to completely block young people from entering the labour market at or before matric. So, the 400 000 annual funnel will be blocked. In addition 200 000 will be mopped up on an annual basis among the unemployed.

The impact of this intervention will be to shift the curve to the left to in Figure 2. The implication will be a drastic fall in youth unemployment from A to B, i.e. from to. However, if the youth have bargaining power, then the unemployment rate falls to and the youth wage rate will increase to. This will generate a living wage and improve income distribution, as the reserve army of the unemployed youth taken back into education and training facilities.

There are creative ways in which the FET sector, and indeed the post-school education system can be ramped up. We may think of recruiting teaching and training staff from other countries, as we build domestic capacity to do training ourselves. We can mobilise existing professionals and those that have retired to provide the necessary training and expertise. Institutions can share resources and optimise on scale economies through for example, combing examinations and testing.

On the other hand, young people must be encouraged to register and seriously take their studies through a stipend. Therefore, the R5 billion that was budgeted for the youth wage subsidy can be used to supplement a national stipend fund to encourage targeted young people to pursue studies and training, whilst the funds that are collected by SETAs through the skills levy can be used to support the public post-schooling and training sector.

Footnotes:

1See Quarterly Labour Force Survey (2009 Quarter 2).

2International Perspective on Youth Unemployment, ILO Presentation to NEDLAC Workshop, 23 January 2012.

3Towards a 15-Year Review, The Presidency, 2009, p.18. R322 a month is said to be a "high poverty line", but a crude calculation shows that this cannot cover items such as cooking oil, soap, sugar, tea, clothes, transport etc. if an individual buys 12.5kg of mealie-meal, 4 full chicken portions and 12 loaves of bread.

4The Monthly Earnings of South Africans (2010), Statistics South Africa.

5The New Growth Path-The Framework, p.3.

6Commission on Employment Equity Report (2009-2010), p.30.

7 Confronting Youth Unemployment, p.14.

8 Ibid, p.34-35.

9Katz L. (1996): Wage Subsidies for the Disadvantaged. NBER Working Paper No.5679.

10Department of Business, Innovation and Skills: Government's Response to Business & Enterprise Select Committee's "The Automotive Industry in the UK" Report, September 2009, p.10.

11Van Reenen J. (2003): Active Labour Market Policies and the British New Deal for the Young Unemployed in Context. NBER Working Paper No. 9576.

12Galasso G., Ravallion M. And Salvia A. (2001). Assisting the Transition from Workfare to Work: A Randomised Experiment. World Bank Policy Working Paper 2738.

13Ibid., p.16.

14Richardson J. (1998): Do Wage Subsidies Enhance Employability? Evidence from Australian Youth. Centre for Economic Performance. London Schools of Economics and Political Science.

15Knight G. (2002). An Evaluation of the Australian Youth Wage Subsidy Special Youth Employment Training Program, SYETP. PhD Dissertation. School of Economics and Political Science. Faculty of Economics and Business. University of Sydney, Australia.

16Dorrance G.S. and Highes M. (1996). Working Youth: Tackling Australian Youth Unemployment. CIS Policy Monographs 34.

17Schaffner J.A. (2001). Turnover and Job Training in Developing and Developed Countries: Evidence from Colombia and the United States. Boston University.

18Attanasios O.P., Kugler A.D. and Meghir C. (2009). Subsidising Vocational Training for Disadvantaged Youth in Developing Countries: Evidence from a Randomised Trial. IZA DP No.4251.

19Attanasios et.al. (2009) mention that "courses included training for: taxi and bus drivers; office assistants; call centre operators; nurses' and physicians' assistants; pharmacy assistants; hairdressing and cosmetology assistants; inventory assistants; archival assistants; pre-school teacher assistants; cashiers; payroll assistants; assistants for computer installation and maintenance; textile operators; wood-cutting machine operators; carpentry assistants, plumber assistants, and electricians' assistants".

20See Schaeffer (2001) in this regard.

21Morales L. (7 April 2011). Columbia's Barriers for Youth Labor Market. Americas Quarterly.

22Betcherman G., Daysal N.M. and Pages C. (2010). Do Wage Subsidies Work? Evidence from Regionally Targeted Subsidies in Turkey. Labour Economics 17, p.710-722.

23Wilson S. and Fretwell D.H. (1999). Public Service Employment: A Review of Programs in Selected OECD Countries and Transition Economies. Social Protection, Labour Markets, Pensions and Social Assistance. World Bank.

24Fretwell D.H., Benus J., and O'Leary C.J. (1999). Evaluating the Impact of Active Labour Market Programs: Results of Cross-country Studies in Europe and Central Asia. World Bank Discussion Paper.

25Leetma R. and Vork A. (2003). Evaluation of Active Labour Market Programmes in Estonia. Draft Manuscript.

26Rotgers G.P. and Arendt J.N. (2010). The Effect of a Wage Subsidy on Subsidised Firms' Ordinary Employment. AKF Working Paper.

27Boockmann B., Zwick T., Ammermuller A. and Maier M. (2007). Do Hiring Subsidies Reduce Unemployment Among the Elderly? Evidence from Two Natural Experiments. Discussion paper No. 07-001.

28O'Leary C.J. (1998). Evaluating Active Labour Market Programs in Poland. Upjohns Technical Report No.98-012.

29Ibid, p. 134.

30Larsson L. (2000). Evaluation of Swedish Youth Labour Market Programmes. Uppsala University and Office of Labour Market Policy Evaluation, Sweden.

31Larsson L. (2003). Evaluation of Swedish Youth Programmes. Journal of Human Resources 38, p.891-927.

32This point is also made by Robert J. Gordon (2009, p.307) in his undergraduate text on Macroeconomics: "Solutions for low skill fall into three categories: better public education, subsidies for firms to train workers, and government-financed training programs. Better public education is essential, particularly for students in disadvantaged backgrounds, since training subsidies and programs will not work if teenagers and young adults cannot read or perform basic arithmetic".

33Gross D.M. (1997): Youth Unemployment and Youth Labour Market Policies in Germany and Canada. Employment and Training Papers 37. International Monetary Fund.

34See van Reenen (2003), p.8.

35ILO (2011). Tackling Youth Employment Challenges: An Overview of Possible Actions and Policy Considerations, p.76.

Issued by COSATU, June 24 2012

Click here to sign up to receive our free daily headline email newsletter