Refugees, Education Outcomes and Gambling for Redemption

Over 80% of refugees live in developing countries – a fact you’d be forgiven forgetting amidst the media frenzy surrounding the refugee crisis in Europe. The infographic below demonstrates this, showing East Africa and the Middle East to be global refugee hotspots: the former due to insurgency in DRC, South Sudan and Somalia; the latter due to the collapse of the Syrian state.

Source: UNHCR statistics

Another oft-neglected fact is that most refugees do not live in camps. UNHCR estimate that globally, less than a 1/3 of refugees reside in camps – the rest live almost exclusively in cities and peri-urban areas. Many refugees relocate to the capital: for example, in Lebanon, over 25% of refugees reside in Beirut.

Source: UNHCR Syria Regional Response data

Refugee Access to Education

Education delivered in refugee camps is imperfect. Curriculums are often not aligned with the system in their home country; the language of instruction is often foreign; teaching materials are scarce, and teachers are in dangerously short supply. This final point is particularly vexing: many refugees are qualified teachers but are barred from formal employment, leading to the surreal situation where there are both not enough teachers and teachers without jobs.

Despite these issues, most refugee children in camps have access to education. This is largely due to the logistical convenience that refugee camps provide: by concentrating refugees in one place, host governments and aid agencies can quickly build tents, distribute food, deliver inoculations and set up rudimentary schools.

By contrast, the average urban refugee is less likely to be in school. Two years ago, I volunteered for Xavier Project, an NGO focused on increasing urban refugee access to education in East Africa. We estimated that in Nairobi, 65% of primary school-aged and just 33% of secondary school-aged children were attending school – making refugee children five times more likely to be out of school than their non-refugee peers. There is significant variation amongst country of origin: in Kampala, we estimated that just 26% of school-age Congolese refugee children were attending school.

There are several reasons for this. Most obviously, refugees from Somalia and DRC often cannot speak English – the primary mode of instruction in both countries – creating a Catch 22 whereby they can’t attend school because they can’t speak English, but they can’t learn English because they can’t attend school. Refugees also face a minefield of bureaucratic obstacles: many can’t enrol in school because they lack identity cards; many more can’t enrol as they lack previous academic transcripts, which in many cases were left behind in a rush to flee violence.

It is not just a lack of education that urban refugees must contend with. They must also tolerate slum dwellings, frequent xenophobia and similarly hard-to-access healthcare facilities. In many cases, it is often only through remarkable social networks that refugees manage to survive.

Faced with this situation, a reasonable question to ask is: why do refugees migrate to the city in the first place? I think I can answer this question, but to do so requires a brief detour to one of the most famous models in Development Economics: the Harris-Todaro model of rural-urban migration.

The Harris-Todaro Model

The puzzle that Harris and Todaro sought to address was this: why do so many people migrate from rural to urban areas in the face of high urban unemployment and poverty? Their answer: calculated risk. In agriculture, workers are assumed to be all-but-guaranteed low-income/subsistence work as a farmer. In the city, there is a high probability they will end up unemployed, living in a slum and forced to eke out a living in the informal sector. But there is also a chance they will land a formal job – with all the salary and non-salary perks that this implies. People migrate until the low-risk, low-reward prospect of agriculture equates with the high-risk, high-reward returns offered in the city.

A numerical example may help illustrate this. In the table below, people are guaranteed an income of 50 in the countryside. If they migrate to the city, there is a 50% chance they will be unemployed (earning an income of 20 in the informal sector) and a 50% chance they land a formal job, which pays 80.

Assuming agents are not risk averse, they should be indifferent between remaining in the countryside and migrating to the city. Subsequently, anything that increases their likelihood of landing a formal job – or increases the salary these jobs command – should incentivise greater rural-urban migration.

Gambling for Redemption

While this model was originally devised to explain rural-urban migration, I think it does a good job explaining camp-urban migration decisions too. For a refugee, the low-risk, low-reward outcome is often remaining in the camp: while they are least virtually guaranteed basic housing, healthcare and education, they are nonetheless prohibited from formal employment, or registering a business. The risks of migrating to the city are extensive and have been described above. But the potential rewards are also great. There is a possibility, however small, that they will obtain formal employment. There is also a chance they will get sponsored by an NGO like Xavier Project, which tend to be disproportionately located in cities. Finally, through being near UNHCR headquarters, they may be better able to ‘play the system’ and make their case heard, maximising their chances of being resettled in a developed country. Hence, camp-urban migration can be considered a similar gamble to rural-urban migration – though one with arguably higher risks and rewards.

This simple model can be applied to risk taking in more extreme scenarios. Consider refugees taking rubber dinghies across the Mediterranean to Europe. This is an extremely high-risk strategy; but one that must be compared to the often-dire alternative of staying put. When framed this way, such a decision seems undeniably risky but not irrational. Rather, the decision simply reflects the brutal reality that refugees must confront.

[974 words]

References

This blog post was largely inspired by my time volunteering for Xavier Project, an NGO based in Kenya and Uganda. They do great work: sponsoring refugee children through school, teaching English to those without and running adult education hubs to teach vocational skills. Please check them out here: http://xavierproject.org/

•           Harris, J. & Todaro, M. (1970), ‘Migration, Unemployment and Development: A Two-Sector Analysis’

Are Humans Getting More Intelligent?

Human beings are getting smarter – at least according to a recent meta-analysis conducted by researchers at Kings College London. They base this claim on the observation, first noticed by James Flynn, that worldwide IQ test performance (the de facto measure of intelligence) has profoundly and consistently increased over the last century. These gains have been virtually universal, with the biggest leaps seen in China and India.

Since IQ tests are always standardised to 100, one way to measure this increase is to have new test subjects take older tests. When Americans today take IQ tests from a century ago, they score an extraordinarily high average IQ of 130. Conversely, Flynn argues, if the American’s of 100 years ago took today’s tests, they would have an average IQ of 70 – the recognised cut-off for people with intellectual disabilities. Progress is most pronounced in the Raven’s Matrices: a non-verbal test designed to measure abstract reasoning (example below). These findings are colloquially known as the ‘Flynn effect’.

Above: US gains in IQ. Source: BBC

Above: Ravens Progressive Matrices, which should be familiar to anyone who sat the 11+.

Test Wiseness

When I first heard of this effect, I was quite sceptical. My initial reaction was that the improvement was most likely due to humans getting better at sitting IQ tests (a phenomenon known as ‘test wiseness’) – not due to their getting more intelligence. For example, in Estonia, when psychologists Olev and Aasa Must laid IQ tests from the 1930s alongside papers from 2006, they found an increase in correct answers – but also incorrect ones. Students had realised they would not be penalised for incorrect answers – they had become better at ‘gaming’ the system.

I no longer think this explanation captures the entire story. Firstly, standardised IQ tests have been an accepted part of American culture for a long time; secondly, the frequency of IQ testing has waned during the last 50 years. In other words, people have ‘gotten used’ to taking IQ tests while simultaneously have had less chance to practice them. Despite this, IQ increases have remained steady over the last 50 years (at about 3 points per decade). Test wiseness alone struggles to explain this fact, which has led theorists to propose several alternative explanations that couch these gains in the vastly different world we inhabit.

Education

For example, one explanation cites the vast increase in the quantity and quality of education as the main reason for the improvement in IQ scores. Globally, more people are in school than ever before. Curriculums have also shifted further towards mathematics and natural sciences, requiring a greater emphasis being placed on classification, logical consistency and abstract reasoning (Blair et al., 2005). This shift away from memorisation towards abstraction has permeated through all levels of schooling. For example, in Ohio in 1910, 14-year olds would have sat state examinations which predominantly tested them for concrete, socially-valuable information, such as the names of all 45 state capitals. In 1990, these same examinations had shifted to almost exclusively abstract questions like “why is the largest city in the state rarely the capital?”. The tenor of education has changed: children are more encouraged to think in the abstract and take the hypothetical seriously – traits we tend to associate with fluid intelligence and higher IQ scores.

Environment

These shifts in education practices have occurred contemporaneously with a radically changing working environment. In 1900, only 3% of Americans worked in professions deemed ‘cognitively demanding’, such as doctors, lawyers or bankers. Today, this figure has risen to 35%. Not only this, but there has been an ‘upgrading’ of these professions: a doctor today must confront a much more complex and information-laden working environment than a doctor a century ago.

The world around us has also changed. Society asks us to solve a wider range of cognitive problems and process a much larger amount of information than before. To deal with this, we have developed mental models to categorise, analyse and make sense of an increasingly chaotic world. These models are often faulty: but our sheer frequency of using them stretches our cognitive faculties to a degree incomparable with a century ago.

One of the most striking ways the world has changed is in it becoming more visual. From pictures on the wall to movies to television to video games, each successive generation has been exposed to far richer optical displays than the one before. Beyond merely looking at pictures, we also analyse them. Picture puzzles, mazes, exploded views and complex montages appear everywhere—on cereal boxes, on McDonald's wrappers, in the instructions for assembling toys and in books intended to help children pass the time. This may have helped us develop a skill deemed an important component of intelligence: visual reasoning. Indeed, the fact that the greatest IQ gains can be seen in Ravens Progressive Matrices (which predominantly test for visual reasoning) may stand testament to the importance of this trend.

Humans 2.0

What to make of all this? It is important not to get carried away. A human brain today is indistinguishable from a human brain a century ago[1]: 100 years is far too brief a period for evolutionary processes to have taken place - nature has not upgraded us to a smarter breed of human. Instead, if we are to take the Flynn effect seriously, it makes more sense to view it as a case of nurture over nature. Consider two identical twins, one of whom becomes a professional athlete while the other becomes an investment banker. At autopsy these twins would look physically very different, despite their similar physiology at birth. In the same way, a human born today will have to endure far more rigorous cognitive training than one born a century ago. While they may have had similar brains at birth, comparing their brains during middle age will likely reveal significant discrepancies in a range of cognitive functions – discrepancies which reflect the different demands placed on them by the different worlds they inhabit.

[996 words]

References

For anyone interested in this topic, I would recommend James Flynn’s Ted Talk on why our IQ levels are higher than our grandparents. Blair et al. (2005) provide a compelling case for curriculum change being the primary driver of the increase in test scores. The BBC News article also provides a nice overview.

•          BBC News, ‘Are humans getting cleverer?’ https://www.bbc.com/news/magazine-31556802
•          Blair et al. (2005), ‘Rising mean IQ: Cognitive demand of mathematics education for young children, population exposure to formal schooling and the neurobiology of the prefrontal cortex’
•         Flynn, J. ‘Why our IQ levels are higher than our grandparents’ https://www.youtube.com/watch?v=9vpqilhW9uI&t=36s
•          Kings College London, ‘A Cross-Temporal Meta-Analysis of Raven’s Progressive Matrices: Age groups and developing versus developed countries’

---

[1] One challenge to this view stems from the observation that human beings are taller today than a century ago, having grown at a rate of 1cm per decade. It is not inconceivable that this increase in height has led to a corresponding increase in brain size. Some scientists, such as Richard Lynn, argue that this could account for the entirety of the Flynn effect. I omit a full discussion of this explanation, as the evidence for the link between height and brain size and between brain size and intelligence is inconclusive.

Why Does Education Reform Happen Where and When It Does?

Education reform fundamentally changes the quality, composition and demands of a workforce. It does so in a way that does not benefit all sections of society equally. This post attempts to outline a simple model to help us think about the winners and losers of education reform, in order to explain why reform happens when and where it does. Doing so will reveal the curious relationship between land reform and education reform, which offers a clue as to why the timing and extent of reform differs so greatly between countries.

A Model of Education Reform

In economic models, we often make crude abstractions which simplify reality. We do this to make our analysis workable and to remove irrelevant details which risk obscuring the main message we wish to convey. These abstractions are often criticised as being ‘unrealistic’. However, the relevant criterion is not whether such abstractions are realistic but whether they are useful. Maps provide a good analogy to this: a map which detailed every single crack, pothole and tree along a road would be realistic but ultimately useless. By contrast, an ‘unrealistic’ map which only detailed the most salient features of a road (such as the name, type and direction the road is headed) would be far more useful.

In the same way, ‘unrealistic’ Economic models can nonetheless greatly aid our understanding of the world. For example, in the context of 19^th century Britain, a crude (but useful) abstraction to illustrate the timing of education reform is to categorise people into three separate groups: landowners, capitalists and workers. Landowners own land (e.g. farms), capitalists own capital (e.g. factories) and workers own their labour time, which they can sell to either landowners or capitalists. In this model, both landowners and capitalists make up the political elite, who must decide on a variety of public sector measures including education policy.

Above: A cartoon parodying the dynamic between the three actors in our model. Source: Henry Devon George Society

Crucial to the timing of reform is the idea that landowners and capitalists demand fundamentally different things from their workforce. Crudely, landowners desire a workforce that is predominantly rural and immobile, so they can extract rent from tenants who have little in the way of ‘outside options’ to recourse to. On the other hand, capitalists desire a workforce that is educated, so that there are engineers to fix their machines, architects to build their factories and at the very least literate factory workers who are able to read instructions. Faced with these contrasting incentives, the pace and timing of reform is dictated by the political clout of capitalists relative to landowners at any one moment in time. When landowners are more powerful, education reform is likely to stall, as such a policy would likely require tax increases and threaten the immobility of their workforce.

To tie this back to land ownership, consider that the political salience of the landlord class is more pronounced when land ownership is more concentrated in the hands of an elite group: the landed aristocracy. Thus, land reforms which equalise land ownership will reduce the power of landlords relative to capitalists, increasing the likelihood for education reform to take place. Not only do they make landlords less important as a political class, but they can may also prompt landowners to diversify their portfolio away from land and towards capital, lowering their economic incentive to block education reform.

Historical Evidence

History provides us with many instances of this sequence taking place. For example, in Japan in 1871, an Imperial Decree initiated the abolishment of the feudal system. Decisions on land utilisation and choice crops were transferred from landlords to farmers and prohibitions on the sale of farmland were removed. These reforms had the effect of greatly reducing the power of Japan’s landed aristocracy. Education reform soon followed: in 1872, the Education Code established compulsory and locally funded education. While in 1873 only 28% of school-age children attended schools, this ratio increased to 51% by 1883 and 94% by 1903 (Gubbins, 1973).

This pattern was mirrored in Russia a few decades later. As the Tsar’s grip on power weakened during the early 1900s, the political power of the wealthy landowners gradually declined leading to a series of agrarian reforms initiated by the Stolypin in 1906. Restrictions on the mobility of peasants were abolished, fragmented land-holdings were consolidated, and the formation of individually owned farms was encouraged and supported through the provision of government credit. As a result, the land holdings of the aristocracy declined from about 35-45% in 1860 to 17% in 1917. As their political power correspondingly weakened, the Duma began to initiate a series of education reforms, leading to the education share of the Provincial Council’s budget increasing from 20.4% in 1905 to 31.1% in 1914 (Florinsky, 1961).

Interestingly, this converse relationship between the concentration of land ownership and the extent of education delivery remains visible today. In Costa Rica and Colombia, where coffee is typically grown in small farms, education expenditure and schooling outcomes are significantly higher than in Guatemala and El Salvador, where large plantations still dominate.

From this simple model of landowners, capitalists and workers, we can derive several interesting conclusions. First, anything that reduces the political power of the landowner class should speed up education reform. Land reform is one way to do this, though certainly not the only way. Electoral reform and a gradual diversification away from agriculture should both have similar effects. Second, in countries where the boundary between landowners and capitalists is more fluid, this should increase the likelihood of education reform, as it enables landowners to diversify away from land in response to political change. Finally, the model implies that while land abundance may initially be a positive thing, it may eventually hamper education reform (and ultimately growth) to the extent that it intensifies the political power of landowners. This may help explain what Acemoglu and Robinson refer to as the ‘reversal of fortune’: the observation that many historically rich civilisations (e.g. the Aztecs) have moved towards the bottom of the income distribution since the industrial revolution.

[999 words]

References

The model outlined in this article was devised by Galor et al. (2009).

•         Florinsky (1961), ‘Encyclopaedia of Russia and the Soviet Union’
•     Galor et al. (2009), ‘Inequality in Land Ownership, the Emergence of Human-Capital Promoting Institutions, and the Great Divergence’
•     Gubbins (1973), ‘The Making of Modern Japan’

Education as a Tool to Build Nations

Imagined Communities

What is a nation? Defining this is harder than it sounds. Nations are not physical objects like trees or mountains. They also seem to be more than the sum of their observable parts. You can conquer a nations’ land, envelop its people, burn its flag and even revoke its legal status without extinguishing it entirely – just ask the people of Kurdistan, Quebec or Palestine.

Nations can be effectively viewed as imagined communities. Though members of even the smallest nation will never know most of their fellowship members, in the minds of each lives the image of their communion. This image can be exceptionally powerful: it can send people to war, paper over societal divisions and help coordinate people en masse.

As a result, nations are malleable constructs that change when the stories that underpin them change as well. This can happen remarkably quickly: for example, the nation of East Germany ceased to exist virtually overnight, as the story behind it was rewritten following the fall of the Berlin Wall. One of the most effective ways to rewrite these stories is through education, via its ability to shape the collective imagination of citizens. In this post, we focus on two comparative case studies to demonstrate this: Tanzania and Kenya and India and Pakistan.

Tanzania and Kenya

At independence, Julius Nyerere’s Tanzania and Jomo Kenyatta’s Kenya were remarkably similar. Both had similar East African geographies, both contained a similarly diverse composition of different ethnic groups and both inherited a similar set of institutions brought about by British colonial rule.

Despite these similarities, the two nations quickly diverged due largely to the disparate personalities of their inaugural leaders. Nyerere was perhaps Africa’s closest approximation to Plato’s ‘philosopher king’: a highly educated visionary committed to the lofty ideals of self-determination, equality and pan-African socialism. These ideals drove Nyerere to pursue an ambitious nation-building programme in which education played a vital role.

Above: Tanzania’s ‘philosopher king’, Julius Nyerere. Source: www.infed.org

For example, via changes to the curriculum, Swahili was institutionalised as the lingua franca of Tanzania. Even students who entered primary school with no knowledge of Swahili soon learned and brought it home to their villages, enabling Tanzanians to communicate with each other in a common tongue. The new curriculum also placed heavy emphasis on ‘civic education’: a subject designed to spread stories about shared Tanzanian identity, ideology and memory. By the 1970s all teachers were required to serve in the paramilitary national service organisation, helping to indoctrinate them into the ideals of the Tanzanian regime. The net result of this was that Tanzanian schoolchildren left school with a common language and a shared idea of their past, present and destiny as fellow Tanzanian citizens.

Across the border, Kenyatta was driven less by ideology and more by the harsh realpolitik of maintaining power in a divided society. Instead of building a nation, he sought to entrench the power of his ethnic group (at the expense of others) by rewarding his fellow Kikuyu with Cabinet positions and lucrative business contracts. This lack of nationalist impetus was reflected in education policy: Swahili competed with English and local languages (e.g. Kikuyu) as the primary mode of instruction in school and the official geography, history and civics curriculum did not study Kenya as a nation until grade 5. As a result, many Kenyans left school without a common language and without a shared idea of what it meant to be Kenyan – in stark contrast to what happened in Tanzania.

The consequences of this have been wide-spread. While Tanzania has experienced virtually no ethnic conflict since independence, Kenya has suffered ethnic violence at every election since Kenyatta’s rule. Moreover, as Miguel (2005) shows, public goods provision is significantly lower in Kenya: people are less willing to give resources to help their fellow citizens. Since their ethnic makeup at independence was virtually identical, these differing experiences can be largely traced to greater nation-building efforts in Tanzania – of which education was an essential component.

Interestingly, Tanzanian nation-building efforts were typically not extended to Zanzibar, the semi-autonomous archipelago off the coast of Dar Es Salaam. Little effort was made to infuse a pan-Tanzanian identity into the curriculum, as the Zanzibar Ministry of Education was left (and remains) largely autonomous of the mainland. As a result, the Zanzibari people have generally not ‘bought in’ to the Tanzanian myth, resulting in a powerful secessionist movement which breeds instability and violence at election times.

India and Pakistan

At Independence in 1947, two new nations of India and Pakistan came into being facing ostensibly similar situations. Both were products of the same colonial project with a broadly coherent education policy: Empire sought to educate a ‘pro-British Indian elite’ through a system that was selective rather than universal, liberal rather than scientific and imperial rather than national. However, colonial educational institutions were unevenly distributed and relied on local collaboration, challenging the idea of initial equivalence. At the time of the first Census in 1950, literacy rates between India (20%) and Pakistan (14%) already exhibited a discrepancy.

India and Pakistan pursued similar educational policies at the outset: a broadly secular education communicated through a unifying language (Hindi for India, Urdu for Pakistan). Both projects expressed a prescriptive vision that was not merely postcolonial but anti-colonial: whilst building on colonial institutions, education offered an opportunity for intellectual decolonisation. 

There were differences, which reflected the different national challenges each country faced. India’s national identity was to be built around a tolerant multiculturalism that explicitly rejected communalism. In recognition of India’s immense linguistic, ethnic and religious diversity, it was hoped that education would serve to enfranchise diverse communities and dismantle identity hierarchies such as caste. Across the border, the ethnic incongruity between East and West made Pakistan an immediately less stable state: the inclusion of Islam in educational policy reflected an urgent need to build a national identity that could make sense of this unique geography.

These initial ideals have not been realised. In addition to failures of policy implementation, both visions of national identity have undergone change almost beyond recognition. In 1977, Zia-ul-Haq launched a coup in Pakistan and sought to legitimise his dictatorship through an agenda of Islamization. As with Jinnah’s vision thirty years previously, education policy was mobilised to express a new national identity. The right to education, contents of curriculum and its purpose were all realigned to the values of the Quran and Sunnah, while national identity was re-expressed along lines of solidarity with the Sunni Middle East, at the expense of cohesion within Pakistan.

In India, Nehru's educational model of secularism and internationalism was consistently rejected by the Bharat Janata Party, whose advocacy of Hindutva, Hindu majoritarianism, became part of educational policy with the advent of their parliamentary majority in 1998. The return of Congress to power in 2004, followed by the subsequent resurgence of the BJP under Narendra Modi, has led to a certain flip-flopping of educational policy: entire textbooks are rewritten according to the ideological bent of the incumbent political party. At government schools, Indian children are taught historical fantasies as facts, as Hindutva legitimises and enforces a national identity that is explicitly anti-Muslim and self-reliant.

From broadly similar starting points, India and Pakistan pursued similar policies at Independence, both of which were undermined by ideological re-conceptualizations of nationhood.  Whilst neither nation has achieved the success seen in East Asia, India’s educational progress has continued to outstrip its neighbour’s. To explain this, we might look to the fact that the ideological reconceptualization of education policy manifested itself in India two decades later than Pakistan. We may also return to the unique strains on Pakistan’s nationhood in 1947 and question whether, as Sen cautiously notes, ‘just as education influences culture, so can antecedent culture have an effect on educational policies’.

[1,296 words]

References

The idea of viewing nations as imagined communities is most commonly attributed to Benedict Anderson. Yuval Harari presents this idea in a more contemporary and accessible fashion, arguing that our ability to believe in ‘imagined communities’ differentiates us from other species in the animal kingdom, and has enabled us to evolve from insignificant apes in the Horn of Africa to the dominant species on Earth.

The Tanzania and Kenya section draws heavily on the work of Ted Miguel, whose work on public good provision helps make tangible the impact of the different commitments to nation-building between the two countries.

The idea of India as an imagined community is best articulated by Sunil Khilnani. Amartya Sen’s coverage of the Subcontinent and development policy is essential reading, not least due to the author’s own experience of Partition. He argues that the current Indian government is repeating some of Pakistan’s mistakes of nation-building, particularly regarding ideological reworkings of nationhood.

•          Anderson. B, ‘Imagined Communities’
•          Harari. Y, ‘Sapiens’
•          Miguel. E, ‘Tribe or Nation? Nation Building and Public Goods in Kenya versus Tanzania’
•          Khilnani. S, ‘The Idea of India’
•          Sen. A, ‘Development as Freedom’

Improving Education Outcomes: What We've Learnt from Randomised Evaluations

In my last post, I discussed the use of Randomised Controlled Trials (RCTs) in Education Economics, focusing on the issues of causal identification they seek to address. If you are unfamiliar with RCTs and have not read that post yet, I would recommend you do so [here].

In this post, I discuss what we have learnt from 20 years of randomised evaluations. I argue that effective interventions (i.e. ones that improve attendance or exam scores) typically do two things well: first, they narrow their focus onto a specific malfunction that is inhibiting education outcomes; second, they do so in a way that properly accounts for the incentives of the agents studied.

What Works

Conditional Cash Transfers

Conditional cash transfers (CCTs) provide a good example of this. CCTs are remarkably simple: participants in the treatment group receive a direct cash transfer (usually ~$50 per month) if their child attains a certain level of attendance in school. CCTs work well as they directly incentivise parents to send their children to school while simultaneously relieving a financial constraint that may have previously prevented them from doing so. In a review article, Glewwe and Muralidharan (2015) outline 27 RCTs to test CCTs, 24 of which produce a statistically significant increase in school attendance. As argued in my last post, the unanimity of these results in a variety of contexts makes the effectiveness of CCTs an extremely robust finding. Of course, CCTs cost money – but the financing could theoretically be provided by donor funding, particularly if it is redirected away from other forms of budgetary support with less proven results.  

Teacher Performance Pay

Another good example is the use of teacher performance pay. Again, performance pay seeks to address a specific malfunction (demotivated teachers) in a way that is appreciative of their incentives. In the same review article, Glewwe and Muralidharan cite 5 high-quality RCTs, all of which show that exam results significantly improve in treatment schools where teachers are paid according to performance. Not only this: Kremer et al. (2005) show that there is a significant positive correlation between teacher performance and the extent of performance pay desired beforehand, suggesting that good teachers ‘know who they are’, implying additional sorting benefits from performance pay (i.e. better quality teachers are attracted in the first place).

What Doesn’t

Discretionary Rewards

However, performance-based interventions fail when insufficient attention is paid to incentives. A good example of this is a study by Kremer and Chen (2001), which gave Kenyan headteachers discretion to award bicycles to pre-school teachers based on good attendance. Somewhat predictably, every teacher was reported to have good attendance and received a bicycle. However, when absence was externally audited, attendance had not changed at all from baseline. The study did not take into account the headmaster’s incentive to collude with his fellow teachers.

CCT programmes must also be mindful of incentives. Suppose you are a farmer with four children who help out with farm work when required. An NGO comes along and offers you money if you send two of your kids to school – how might you respond? A sensible way might be to remove your non-sponsored kids from school completely, to compensate for the shortfall of labour on the farm. Several studies have documented these types of ‘negative spillovers’ to siblings (e.g. Barrera-Osorio et al., 2011); interventions must be wary of these and redesigned accordingly.

More-of-the-same Inputs

Interventions may also fail when inadequate care is taken to understand the fundamental barriers to learning. For example, when it comes to improving test scores, ‘more-of-the-same’ inputs (e.g. more textbooks, stationary and teachers) have proven remarkably ineffective (Kremer et al., 2013). Buying more textbooks may sound sensible, but when children are illiterate, unsupervised and of wildly varying abilities, these textbooks are unlikely to do anything but gather dust. These examples should caution us against ‘lazy’ interventions that throw more of the same ingredients at a problem and expect different results. We need to be more creative. A study by Banerjee et al. (2007) sets a good example: to deal with the problem of ‘left behind’ children, the authors implement and test two novel remedial education programmes. The first hired young women from the community to teach basic literacy and numeracy skills; the second consisted of a computer-assisted learning programme tailored to the child’s ability. On average, the two interventions increased test scores by 0.28 and 0.47 standard deviations respectively.[1] The author’s estimate that extending this programme would be 12-16 times more cost effective than simply hiring more teachers.

The Issue of Scalability

While these are undoubtedly promising, there are numerous difficulties associated with scaling up these interventions to form meaningful policy change. For instance, RCT-based interventions may change market prices or produce negative spillovers when scaled up, both of which may shift the net benefit of the programme or the identities of the beneficiaries. On a more practical level, wide-scale policy change often requires immense political will and capacity – neither of which are required for a successful RCT (Banerjee et al., 2017).

However, these issues are not insurmountable. The experience of Pratham, an Indian NGO, gives cause for hope. In 2004, Pratham ran a successful RCT which grouped children based by ability rather than ages, helping children that had previously been left behind. The Pratham-approach was then integrated into some pilot schools. The intervention had to be adapted to be both feasible and financially viable: several variations were retested by further RCTs to determine the most effective. In the end, the best variation was the adding of a ‘Pratham-specific’ hour to the school day. Today, the additional-hour model exists in 107,921 schools across 13 states of India, reaching nearly five million schoolchildren. Five million children are now being taught daily by a strategy proven to help them learn – the welfare implications of this are staggering. These types of interventions make me excited to work in the education sector, and help convince me that RCTs can be a powerful tool to improve the world we live in.

[998 words]

References

Glewwe and Muralidharan (2015) have a very good review article which outlines the education RCTs that have worked vs. those that haven’t. Banerjee et al. (2017) provide one of the best and most up-to-date discussions of the challenges associated with scaling up successful RCTs to policy change. All cited articles are listed below:

·         Banerjee et al. (2017), ‘From Proof of Concept to Scalable Policies: Challenges and Solutions, with Application’

·         Banerjee et al. (2007), ‘Remedying Education: Evidence from Two Randomised Experiments in India’

·         Barrera-Osorio et al. (2011), ‘Improving the Design of Conditional Transfer Programs: Evidence from a Randomised Education Experiment in Colombia’

·         Glewwe and Muralidharan (2015), ‘Improving School Education Outcomes in Developing Countries: Evidence, Knowledge Gaps, and Policy Implications’

·         Kremer et al. (2005), ‘Teacher Absence in India: A Snapshot’

·         Kremer and Chen (2001), ‘Interim Report on a Teacher Attendance Incentive Program in Kenya’

---

[1] Standard deviation is a measure of how spread out numbers are. Different contexts have very different data distributions and averages; reporting results in standard deviations is a way to ‘standardise’ this so we can easily compare across contexts

Uncovering Causality I: Randomised Controlled Trials

Causality

For the most part, empirical Economics (or ‘Econometrics’) is concerned with identifying causal relationships in the social world. We often think of causation in terms of the possibility of manipulation. For instance, X can be said to cause Y if when we manipulate X, Y responds in the predicted fashion.

In Economics, we propose causal relationships via hypotheses, which are tested using data and standardised statistical techniques. This deference to an objective scientific method is used to justify Economics’ label as a social ‘science’. It is also used to distinguish Economics from humanities subjects like History, which may speculate causal relationships via theory but make little attempt to validate them in a way that is both objective and generalisable.

This is not to say that Economics is the same as Physics or Chemistry. Of course, physicists and chemists share economists’ commitment to making meaningful causal statements about the world. However, in dealing with more fundamental and discernible laws of nature, natural scientists are able to make near infallible ‘if – then’ statements such as ‘if water reaches 0°C, then it will turn to ice’. Moreover, since their objects of interest are atoms rather than humans, scientists can conduct carefully controlled laboratory experiments to make causal identification easier and more replicable.

In contrast, economists study a messy social world containing a maelstrom of different actors, all of whom interact and react with one another in chaotic and unpredictable ways. For every cause we propose to explain an observed outcome there are myriad other potential causes pushing in the same and opposite directions. Isolating the sole impact of our speculated cause can be very difficult, which makes it a lot harder to make robust causal statements.

Nonetheless, Economists have a variety of tools to help do this. To understand these, a useful starting point is to distinguish between correlation and causation.

Correlation ≠ Causation

Correlation ≠ causation. To see this, suppose I collect data on textbooks per classroom and mean exam scores for 30 schools. The scatterplot might look something like this:

Source: Made up figures

Can we conclude from this that more textbooks cause higher exam scores? No – it may simply be that schools with more textbooks also have lower student-teacher-ratios (STRs), better classrooms and more engaged teachers – all of which could also improve exam performance. We need some way to control for these confounding factors.

Randomised Controlled Trials

Perhaps the simplest way to do this is to conduct a Randomised Controlled Trial (RCT). The basic idea is to take a sample of schools and randomly split them into a treatment and control group. We split the groups randomly to help ensure they are ‘balanced’, meaning the treatment and control group are as similar as possible in terms of STRS, desks, and any other confounding variables we can observe. If the two groups are imbalanced, we can always re-randomise until balance is achieved.

The treatment group is then ‘treated’ while the control group is left untreated (or given a placebo). In this instance, a treatment may be to provide schools with free textbooks for each child. After giving the treatment time to take effect, we then examine children in both groups to test for a ‘treatment effect’. The outcome of interest is the difference in exam score between groups: since the two groups were as similar as possible ex ante, we can assume any difference in exam performance post-treatment is caused by the treatment itself.

The diagram below may help conceptualise this:

Source: Sydney Morning Herald http://www.smh.com.au/

This is an RCT in its simplest form. We can add twists to this basic framework: for example, if we are worried about baseline imbalance, we can examine both groups before the intervention and then again at the end of the treatment window. We then test to see if there is a ‘difference-in-difference’ between the two groups, which effectively nets out any ex ante differences between the two.

RCTs take their intellectual heritage from the field of medical science. When testing a new drug, pharmaceutical companies compare the outcomes of a treatment vs. a control group made to be as similar as possible ex ante. While the first published RCT in medicine took place in 1948, it took until the 1990s for RCTs to become popular in Economics. In just 20 years, RCTs have gone from being a novel econometric technique to being considered the ‘gold standard’ in conducting causal work in social science.

Strengths

There are a number of reasons for this. Most importantly, by controlling for confounding factors at the outset, we get arguably the ‘cleanest’ identification of a causal relationship between X and Y. For me, this is where Economics most lives up to its social science moniker. If we conduct the same RCT in a variety of contexts and get the same result each time, this, I think, is the closest we can get to objective scientific knowledge about the social world. Owing to the simplicity of their methodology, the results from RCTs are also easy to communicate, aiding effective policymaking.

Limitations

However, RCTs are also expensive, time-consuming and often raise a host of ethical questions. Also, as my friend Hannah comments on my first post, there have been difficulties in scaling up the findings from RCTs to meaningful policy change – an issue I discuss further in my next post. These points are all important; but for me the most compelling limitation of RCTs is their inability to answer the really big questions. For example, consider the claim that education causes people to earn higher wages. We cannot, practically or morally speaking, randomly allocate different amounts of education to children. In the same way, we cannot uncover the effect of education on growth by randomising education delivery across different countries. To this extent, it is best to view RCTs as a powerful tool with a well-defined purpose, rather than a ‘silver bullet’ which solves all the issues of causal identification.

[990 words]

References

To learn more about RCTs, a good place to start is ‘Running Randomised Evaluations: A Practical Guide’ by Rachel Glennerster and Kudzai Takavarasha.

In my opinion, there is a gap in the market for a book that explains the principles of econometrics in an intuitive and engaging fashion (the best I can think of is Mostly Harmless Econometrics, but even that gets quite technical). If you know of any, please let me know.

A Despots Guide to Education

In the past, student activism has been a force to be reckoned with. It helped end Apartheid in South Africa, was the catalyst in the end of the Vietnam War and was the loudest voice in the pro-democracy protests in China in 1989, epitomised by the infamous ‘Tank Man’ protestor of Tiananmen Square. In Malawi and Benin, students at the University of Malawi and Cotonou University led the first anti-government demonstrations in over three decades, hastening the process of democratisation in Africa (Posner, 1995).

Above: ‘Tank Man’ of the Tiananmen Square protests. Source: www.history.com

As these examples show, education can be a powerful force of history. Traditionally, dictators have not sat idly by for this force to determine their fate; on the contrary, they have actively manipulated this force as a means to serve their own ends. To understand how, it makes more sense to view dictators as skilful political agents who deploy a variety tactics to ensure their survival, rather than blundering, Sasha Baron Cohen-style characters.[1] Maintaining power in an autocracy can be viewed as a delicate balancing act: too much repression and dictators risk inciting a revolution; too much liberalisation and they risk losing their iron-grip on power. To keep this balance alive, dictators have used education policy as a vehicle to employ a variety of strongman tactics, including repression, patronage and indoctrination.

Strategy 1: Repression

One way to deal with the threat students pose is to simply intimidate, arrest and even murder prominent activists. A notorious example of this is the case of Steve Biko, the leader of the South African Students’ Organisation. In 1977, Biko was arrested and brutally murdered by the Apartheid regime. To a lesser degree, we can see this tactic in use today in Uganda. President Museveni has actively targeted popular student activists like Kizza Bisigye and consistently shut down peaceful student protests as soon as they start, with the ringleaders arrested and often given lengthy prison sentences (Parkes, 2015).

While this approach is arguably the simplest, an overreliance on heavy-handed repression can produce martyrs and intensify existing grievances. A more subtle form of repression is to simply slash funding to the education sector. This tactic is taken straight from the colonial handbook. For example, in the DRC, Belgian occupiers capped education at the primary level in order to present the emergence of an aspirational and potentially dangerous educated class of Congolese. As a result of this policy, the DRC had just 12 University graduates at the time of independence (Bayart, 1989).

Strategy 2: Patronage

Dictators also use education as a means to distribute favours (or ‘patronage’) to their support base, whilst simultaneously excluding outsiders. The benefits to this strategy are two-fold: support is consolidated amongst the ‘in’ group whilst the ‘out’ group is gradually disempowered. Patronage is often most effective when it is used to exploit pre-existing divisions in a society, such as appealing to the leaders regional, religious or ethnic group.

Education is a popular vehicle to deploy this type of favouritism. As Franck and Rainer (2009) show in Africa, the education enrolment of a particular ethnic group increases when they have 'their man’ in power. Congo-Brazzaville serves as an interesting case study for this phenomena. The country is divided into two main ethnic groups: the Mbochi in the North and the Kongo in the South. From 1960-1968, the country was ruled by ethnic Kongo rulers. This changed in a coup d’état led by Marien Ngouabi in 1968, which installed a series of Mbochi leaders for the next 24 years. The graph below shows that when Kongo leaders ruled the country, the educational completion rate of the Mbochi were significantly lower than that of the Kongo. However, after Ngouabi’s coup, the Mbochi quickly closed this gap and eventually achieved higher rates of primary school completion (and attendance, and female literacy).

Source: Franck and Rainer (2009)

Strategy 3: Indoctrination

Finally, dictators can use education as a means of indoctrination. Similar to patronage, this strategy does more than just allay the threat education can pose. By inculcating in the youth attitudes and beliefs conducive to the ruling party, indoctrination serves as a means to entrench and solidify the regime for many generations to come.

North Korea’s Kim Jong-Un serves as the undisputed poster boy for this tactic. Amidst stories of the glorious defeat of American invaders, North Korean textbooks are also reputedly littered with bizarre claims about the leader himself. Some textbooks claim that Kim Jong-Un learned to drive at age 3, won a yacht race at age 9 and scored 11 hole in ones on his first ever round of golf, only to promptly retire and to never play the sport again (Daily Mirror).

Above: Pages of a North Korea elementary school textbook. Source: http://blog.jinbo.net/

Conclusion

These examples reemphasise the idea that dictators are not simply insensitive thugs who got into power by accident. More often than not, dictators represent rational political actors versed in using the full extent of the state apparatus to ensure the survival of their regime.

Viewing dictatorship in this light confers a number of advantages. First, it guards us against the complacent idea that dictatorships will inevitably unravel due to incompetence at the top. This idea is supported by the history books, which show dictatorship as one of the most durable forms of government that humans have ever invented. Second, this rational actor framework helps explain many things beyond education policy. For example, it can explain why dictators tend to direct a disproportionate amount of public goods towards cities rather than countryside: a phenomena known as ‘urban bias’. Put simply, rioting factory workers pose more of a threat to a regime than disorganised farmers, so are placated with an unequal share of state resources.

[994 words]

References

Thanks for this blog must go to Brian Klaas, my African Politics tutor at Oxford, who encouraged me to ‘see through the eyes of an elite’ when thinking about autocracy in Africa. Brian has a great book called ‘The Despots Accomplice’ which I would recommend to anyone interested in authoritarianism or democratisation around the globe. 

I would also recommend my friend Charlie Parkes’ article in the Huffington Post, which looks at the strongman tactics being deployed by President Museveni in Uganda: https://www.huffingtonpost.co.uk/charles-parkes/uganda_b_8085226.html

Other articles referenced are given below:

•          Bayart (1989), ‘The State in Africa: The Politics of the Belly’
•         Daily Mirror: https://www.mirror.co.uk/news/world-news/north-korean-school-textbook-claims-5487598
•          Franck and Rainer (2009), ‘Does the Leader’s Ethnicity Matter? Ethnic Favouritism, Education and Health in Sub-Saharan Africa
•          Posner (1995), ‘Malawi’s New Dawn’

---

[1] There are of course exceptions to this: most notable is perhaps Jean-Bédel Bokassa of the Central African Republic, who reputedly viewed himself as the genuine reincarnation of Napoleon Bonaparte. His coronation alone cost 10% of CAR GDP.