Peer Effects in Education in South Africa

A great paper presented at the NEUDC conference by compatriot Rob Garlick on peer effects in education.

Children’s success in school depends, partly, on the actions of their peers. An important policy question is therefore the assignment of children in schools. Should we, for example, track students by separating the “weak” and “strong” students into different classes? Garlick finds that, at least when it comes to dormitories at the University of Cape Town, tracking leads to a reduction in average grades.

Garlick stumbled into a natural experiment at his old University when residential admissions changed from tracking to random assignment. So, he could take the change in average marks in the dormitories, before and after the policy, and compare this to the whole university. He also looked at how particular students (based on their high school grades) are affected by the policy. Turns out that the weaker students benefit massively from being in a dormitory with stronger students; however, the stronger students don’t seem to be badly affected by the reverse. Combing weak and strong students in a dormitory therefore leads to improved grades.  

What explains this peer effect? It seems like it has more to do with the friends you make, rather than the people you work or live with. The peer effects don’t operate across races. Since South African students still tend to socialise across racial lines, this suggests it only helps to live in a dorm with a strong student that you actually socialise with. Another paper at the same conference on peer effects found similar evidence. In a leading business school in India, students are randomly allocated to a dorm and a classroom. The paper finds no peer effects in the classroom, but strong peer effects in the dormitory.

So, what conclusions can be reached from this paper? This is strong evidence that tracking within a school is seriously detrimental to the weaker students. The more interesting question in South Africa, however, is the distribution of education outcomes between schools. Certainly, township schools suffer when their smartest students get scholarships and leave for good schools. But what about the flip side: will students in under-performing schools benefit if they are placed in well-performing schools? This paper suggests that positive peer effects might not operate if students don’t socialise across racial groups. 

External Validity and False Positives in Randomised Control Trials

Two presentations at the latest NEUDC grabbed my attention in providing some necessary caveats on the conclusions we can reach from RCT's. 

First, a paper looked at the scalability of a proven intervention. Researchers intentionally select able NGO’s to implement the projects. (They often also put a lot of individual effort in to ensure quality implementation). However, policy conclusions often involve large-scale government roll out. Can a government repeat on scale the success of a highly motivated and able NGO?

This paper looked specifically at the use of contract teachers in education. (a great summary of this on the CSAE blog). Duflo Dupas and Kremer (2009) show that the use of contract teachers can significant increase educational outcomes in Kenya, partly because they face stronger incentives to teach well. However, turns out that implementation relied on a food NGO. When the NGO scaled up the project it worked, but when the government scaled up the project it didn’t.

This places some caveats on the policy conclusions we can reach from many RCT’s.

The second paper applies the standards required from RCT’s in medical trials to economic papers and finds us severely lacking. We have stolen the method of RCT’s from medicine, but we ignored what they have learnt about the shortcomings of RCT’s. This is a glaring gap and I can’t believe that this paper is the first to do it.

Randomisation solves endogeneity problems; however, biases emerge in the way that we conduct and report our studies, which could lead to false positives.

One big source of bias is lack of “blinding”. Participants respond or act differently because they know they are being treated, a kind of “Hawthorne effect”. This change in behaviour could have nothing to do with the actual treatment. In medical trials this is solved by giving the control group a placebo, but this is far more difficult in social projects. Furthermore, data collectors might ask questions differently in treatment units, because of the perceived pressure from the researcher to get a positive result.

The big problem, of course, is that researchers are biased. Aspirational graduate students (like myself!) invest years in a project and future job prospects often depend on finding a positive result. So, the more discretion is left to the researcher (in sample selection or reporting of results, for example), the higher the bias.

The authors propose introducing standards for conducting RCT’s and reporting results, similar to that in medical trials. The more we can tie the hands of the researchers, the less chance that /she can bias results. This would be a massive contribution to the field.

Marikana and Collective Bargaining - good or bad?

A lot has been written on the role of collective bargaining in the current labour relations in South Africa. How did it impact the Marikana massacre? What are the consequences for the future of collective bargaining? Is this a good or a bad thing?

Collective bargaining means that bargaining agreements are binding at an industry level, rather than an enterprise level.

Some, like sociologist Peter Alexander, believe that collective bargaining is the cause of current labour unrest in South Africa, since it gives undue power to the dominant labour union. If NUM membership has 50%+1 of all union membership in an industry, it can bargain on behalf of all workers. Smaller unions feel unrepresented and not all worker concerns are addressed.  It has the further danger that the dominant union can get too “cozy” with the capitalists, since it is insulated from enterprise-specific labour issues.  

What I don’t understand about this argument is that collective bargaining is weaker in the platinum industry than in the rest of the mining sector; yet it is in the platinum industry where labour conflict is most severe. In all other minerals, agreements are binding for all mines, whereas in the platinum industry, agreements are only binding to other platinum mines. Furthermore, the response of NUM and industry after Marikana, was to strengthen collective bargaining in the platinum industry.

Others, like Jay Naidoo, bemoan the potential demise of collective bargaining, and claims that it is the “cornerstone of our democracy”.

So, is collective bargaining a good thing or not? The evidence seems to point that it is a bad thing for South Africa: bad for employment and often bad for employer-worker relations.

 A recent paper by Berkeley economist finds that collective bargaining in South Africa comes at a prohibitive cost for small firms, and thus contributes to lower employment. South Africa has a smaller proportion of small firms relative to other countries. This is worrying since small firms are a large source of employment in a country. The problem is that small firms have lowe capacity and thus cannot afford the collective agreements, that are commonly reached between the largest firms and unions in an industry. More flexibility for firm-specific labour negotiations would thus allow for more entrants of smaller firms in an indsustry.

Labour mediator Brand claims that collective bargaining often hurts both employer and employee, since it does not allow bargaining on firm-specific issues. He cites the case where Impala Platinum actually wanted to pay a wage premium for Rock-Drill Operators (RDO’s) because of competition from other mines, which NUM rejected.  

The puzzle is that collective bargaining seems to work in other countries. It is seen as one of the reasons for wage equality and stability in north European countries like Germany and Sweden. But then again, the largest source of inequality in South Africa is not wages, but employment. 

Does South Africa need "decent jobs" or just plain jobs?

A priority in South African policy is the creation of “decent jobs”. With an unemployment rate of close to 25% (or closer to 40% depending on how you count), can we as a country afford to focus on “decent jobs”, rather than just plain jobs?

In a recent blog Professor Teal, from Oxford University, he argues that “it is the creation of these kinds of jobs which increases poverty, precisely because such jobs are so scarce”.

Teal attributes the rapid decrease in poverty in Ghana over the past two decades to a dramatic increase in employment, especially in low paying jobs in small firms. He contrasts this with the dual labour market in South Africa where the few who are lucky enough to be permanently employed earn well.

Research in South Africa suggests fact that “decent jobs” would come at a cost of fewer jobs. For example, in a recent paper, Bhorat et al find a significant reduction in employment figures in the agricultural sector in South Africa, due to an increase in the minimum wage.

The poorest in South Africa are those without permanent employment. So, if we as a society care most about the poorest shouldn't the priority be to get more into the formal labour market?

The issue, of course, is that the South African labour market is fundamentally different to other sub-Saharan African countries. We have a highly developed economy floating in a (much larger) underdeveloped economy. This has two implications. First, some people are earning exceptionally well. Large wage disparities breed discontent. Relative deprivation matters for well-being. So, even if the creation of “bad jobs” will do more to alleviate poverty in South Africa, it might not be politically feasible in a highly unequal society. Second, there are more opportunities for “decent jobs” in South Africa, compared to other sub-Saharan African countries. The business community complains of a massive skills shortage. So, if only our education system could improve, then maybe there would be no need to legislate “decent jobs”. 

Service Delivery Protests and Community Monitoring

I believe there is great potential for more community involvement to assure quality provision of public service in South Africa. Just look at the increasing number of public service protests all over the country. People are angry and are prepared to act. However, burning cars and throwing bricks might get government’s attention, but won’t increase nurse or teacher attendance rates. We have seen NGO’s that hold central government accountable (such as Equal Education that pressurize government spending on education), but we also need NGO’s that help citizens hold service providers accountable at a local level. 

What is needed is a more structured way to direct this growing discontent in a productive way to encourage improved service delivery.

There is growing research in economics on how community involvement can lead to improved service delivery. In health care project in Uganda, “report cards” were created on the quality of health care provision in communities. NGO’s then visited these communities, where (i) the report cards were disseminated and discussed; (ii) a “contract” was reached between community members and health care providers on how to improve health care provision; and (iii) measurable deliverables were agreed upon and a system was set up for community members to measure these deliverables. As a result of the project, health care treatment improved dramatically, leading to notable improvements in health outcomes (such as lower child mortality).

Primary Health Care and Community Monitoring

In Jonny Steinberg’s book on HIV/AIDS, “Three Letter Plague”, he follows the life of one young man who refuses to get tested. His life is contrasted with a project initiated by Medecins Sans Frontiere (MSF) and its champion, Herman Reuter, who believes that access to health care is the core problem, rather stigma over the disease. 

What I learnt most from the book was the work from the MSF project.

First, there should be a refocus of health care policy towards primary care. South African health care is vastly under-staffed and under-resourced. However, you don’t need to a doctor to diagnose and treat TB or HIV; you need a nurse or a trained volunteer. Furthermore a nurse is worth more in a rural clinic than in a hospital. The poor living in rural areas do not have access to hospitals. They can only afford to visit a hospital when they are very sick, by which point it is often too late. A lot of prevention and care could be done in clinics.

Second, there is great capacity for communities in South Africa to play a more active role in demanding good quality health care. In the health care project mentioned in the book, MSF initiated HIV/AIDS support groups. These support groups not only provided each other with support, but placed pressure on clinics to supply medicine and provide quality care. More on this, in my next blog

Do Trade Unions in South Africa Represent the Poor?

As trade union members received a pay rise in Marikana, Lomnin closed down one of their shafts, terminating an agreement which affected 12000 casual workers. Casual workers form an increasingly large component of South Africa’s labour force. They receive fewer benefits, are paid significantly less, and have less job security. They are not represented by trade unions. Yet, the actions of the unions directly affect them. Are the casual labourers being screwed by the unions?

Andrew Kerr published a recent working paper on pay differentials in South Africa. Andy, please forgive me for only discussing some basic descriptive statistics (which I concede are subject to some caveats). According the KIDS data on workers in Natal, it seems that trade unions in South Africa really do represent the economic elite. Only 10% of the total labour force is unionised; and they earn significantly more than the rest of the economy. Furthermore, according to this data, a shocking 40% of workers are categorised as casual. 

The Quarterly Labour Force (averaged over the past 5 quarters) show similar trends. Only 30% of those working are unionised; and only 64% of workers claim to have a permanent contract with their employer. Union members earn significantly more

So, how could the actions of trade unions hurt casual worker? Simply put, a trade union represents only its members, and not all of the work force. It will push for higher wages and larger job security and benefits for their members. This is, of course, very important. However, with higher job security and benefits, firms become reluctant to hire and opt for hiring and rehiring casual labour on a short term basis. So, the unintended consequence is a dual labour market, where a large part of the work force is stuck with no job security or benefits. Those who suffer most are the new entrants into the job market – the youth. 

A recent report by the Centre for Development of Enterprise (CDE), summarises study by Neil Rankin from WITS on the difficulties of South African youth in finding permanent employment.

Service Delivery and Culture of Entitlement

The poor get poor service, not only because of bad provision and roll-out, but because the poor expect too little. 

A doctor friend of mine who moved from South Africa to UK recently remarked on the difference in patient behaviour between poor South Africans and the UK. In the UK, patients are more likely to question and challenge the doctors’ decisions. In South Africa, poor patients stoically accept doctors’ judgement. As a result, doctors in the UK are held more accountable by their patients. 

As another example, a South African friend failed to navigate the public health system when his wife got cancer. He was illiterate with no experience with hospitals. But more than that, he was intimidated. He was unwilling to push to receive the treatment he deserved. As a result, uptake is lower.

These examples show that improved quality and use of public service provisions is not only about management and access. It requires a better appreciation of the poor’s engagement with service providers. Accountability and uptake will be lower if clients expect too little. 

What explains this difference in behaviour? Maybe it is low expectations, learned from a history of bad service delivery. Maybe it is lack of confidence: asymmetric power relationships, which still remain after apartheid ended. Maybe the poor don’t believe they deserve good service. 

Annette Lareau (2011) argues that middle class parents cultivate in their children a sense of “entitlement”. They believe they deserve more and as a result get more. So, poor South Africans don’t have too much of a culture of entitlement, as many claim, they have too little. 

If a sense of entitlement is learned in the middle class household, then maybe public policy can do the same for the poor. Maybe it is possible to shift expectations through appropriate provision of information.  Ray (2003) argues that one poverty trap is a “failure of aspirations”. This could be applied to individual dreams (expectations of personal achievement), but also expectations of service delivery.

Morality and the Markets

Social justice philosopher, Michael Sandel, has gone on a recent tirade against the hegemony of economic thinking in society. In particular he thinks we take market exchange too far – we need to bring morality back into the markets. His arguments are either paternalist, or based on a misunderstanding of economics. 

His first point is that economists only care about efficiency.

“We fall into thinking that economic efficiency—getting goods to those with the greatest willingness and ability to pay for them—defines the common good.”

This is a common misconception of economics. The result that pareto efficiency could be reached with free exchange is a positive result, not a normative statement. Economic theory by no means excludes concerns over equity. It just forces you to be explicit about it. There is, in fact, a whole field of economics which analyses formally how different societies would look, given a desire for equality. To my view the key contribution of economic theory to arguments of social justice is a better understanding of the trade-offs that a social planner faces between efficiency (the size of the pie) and equality (the distribution of the pie).

His second point is that “parties [who exchange] may value the things they exchange in the wrong way”. 

This is patronising. It is very arrogant to think that your valuation of a good is superior to another’s. And even if you are right, the solution would be to convince others of your valuation, rather than limit their freedom to exchange.

His third point is the free choice is not always free. In the market for organs:

“In practice, the sellers of kidneys would likely consist of impoverished people desperate for money to feed their families or educate their children. Their choice to sell would not really be free, but coerced, in effect, by their desperate condition.”

 This is true. But the solution would be to create a society which gives the poor more choice, rather than less. It is abominable if a poor person is forced to sell his/her body out of economic desperation. But limiting that choice makes him/her worst off, not better.

I think there are two legitimate reasons to place restrictions on exchange. First, if one party is unwittingly made worse off from the exchange, because it places limits on a person’s future choices. Take for example drug addiction, debt bondage or human trafficking. Second, if exchange between two parties makes a third person worse off. Take for example pollution. Once again, economists have a lot to say about this.

I have so often come across such straw man criticisms of economics. Why is it that people think that the study of economics comes at expense of morality? Why are we so commonly misunderstood? Maybe our thinking gets “lost in translation” when applied by policy makers. If this is true then yes, I agree with Sandel that a better economics education is required.

Markets and the Environment

Reading Jarrod Diamond’s Collapse makes me think that economists need a better understanding of ecological systems when we think about the appropriate role for government in combating environmental problems. Welfare economics has demonstrated the need for government intervention when information is imperfect or markets are incomplete. A lot of market failures can therefore be overcome by developing property rights and providing appropriate information. 

However, Jarrod Diamond’s Collapse makes me think that there are some unique features of environmental problems that our textbooks don’t address: 

1.       Long lags and Multiple Equilibriums: Environmental degradation, and its resulting economic consequences, is not immediately visible. There are thus long lags in price changes and resultant market responses. However, by the time the market adjusts, it might be too late. This is because multiple equilibriums in ecological systems means that environmental destruction is often a one-way street: it cannot be reversed. Diamond’s example is of deforestation which leads to erosion of all land, but overfishing is good current day example, where fish stocks cannot be replenished to its original level.       

 

2.       Systemic Risk and Long-Term Environmental Cycles: Human beings have short time horizons and we are not good at thinking about the risks of unlikely or catastrophic events. As our societies become increasingly inter-dependent, we benefit from the specialisation and exchange. However, inter-dependence also creates risk. Furthermore, fortuitous circumstances today could be nothing more than a centuries-long upward cycle. By increasing our systemic risk, we become more vulnerable to collapse when environmental circumstances change. Diamond cites the collapse of the Anasazi empire the US South-west where a complex inter-dependent society developed over a couple of centuries of good rainfalls, but imploded when rain become less frequent. 

These examples invoke less optimism in the market to adequately respond to risks of environmental destruction.