New paper: education, urban poverty and migration in Bangladesh and Vietnam

For my second working paper for UNICEF's Office of Research I analysed some data from two recent household surveys focusing on urban poor groups. The first was collected in Dhaka, Bangladesh, in 2008, as part of CREATE (and also used here [pdf] and here [£]). The second was the 2009 Urban Poverty Survey in Ho Chi Minh City and Hanoi, Vietnam. I focused on rural-urban migration as a factor explaining urban children's school attendance and how much their parents spent on their education. In Dhaka I found (among a sample living in informal settlements) that rural-urban migrants, and especially recent arrivals, had lower levels of adult education, fewer assets, and lived in worse housing conditions, than non-migrants. Even with statistical controls in place for these differences, children of migrant households had worse educational outcomes than those from those from native households. In Vietnam a major issue was the hokhau system under which households are officially registered as living in a particular area, with consequences for their rights to service provision. Migrant households who had not managed to obtain local registration in the city were worse off than others, and the gap was bigger among the poorest households.

So an obvious question is, are specific policies or policies needed for migrants distinct from those needed for other poor urban groups? Given that migrants, and especially recent migrants, were among the poorest, they would tend to benefit from programmes (stipends, school meals, alternative schooling) that effectively target urban poor households in general. But there were clearly some areas where migrants had specific needs. As well as lifting explicit barriers like hokhau, what seems to be needed is reform to deal with the mobility of populations and rapid growth of migrant-receiving areas within cities, and to help new migrants settle in the city and get the services they are entitled to.

Cities and children in Tanzania

As a companion piece to last year's State of the World's Children report, UNICEF also produced this report on cities and children in Tanzania. Still predominantly rural and known for post-independence 'villagisation', Tanzania is nevertheless urbanizing rapidly. Many of the themes here are familiar from several other studies of urbanization. As elsewhere, there is difficulty surrounding the classification of rural and urban areas, with some large, heavily populated areas continuing to be classified as rural.

As elsewhere there is an urban advantage for children on average. Urban primary schools are better according to most indicators: teacher/student ratios, expenditure, examination results. Pre-schools are more available in urban areas. But as elsewhere, there are many deprived urban areas where this urban advantage is small or perhaps even reversed. A recent survey of 3 districts of Dar es Salaam  found on average 81 pupils per classroom, and reported that teachers often offered private tuition and pupils reportedly risked being put at a disadvantage if they didn't take it. Household survey evidence suggests that costs are more of an obstacle in urban than in rural areas.

In its education section the report struggles a bit with a lack of concrete information, specifically comparing urban poor, urban rich, rural poor and rural rich and can only exhort that "Accurate information is urgently required..." on education for underprivileged urban students. It draws mainly on recent research from the organization Twaweza and its initiatives Uwazi and Uwezo, the latter a 'citizen movement based' approach to assessing literacy and numeracy in East Africa, inspired by ASER in India. It includes some new analysis of inequalities in test scores. But many of the problems highlighted - such as teaching by rote - are no doubt widespread in both rural and urban areas, and some probably affect middle-class students as well as the poorest, highlighting the need for good research that compares across categories.

As has been argued in other contexts, the report notes that national policy and programme frameworks continue to target mostly rural poverty. It argues that gathering sub-municipal data is therefore a "priority for planners, service providers and communities" and calls for urban partnerships to make Child Friendly Cities. For education it recommends that "efforts in cities should aim to achieve an inclusive, affordable model that could be gradually expanded until the promise of the reform – universal quality education – is fulfilled". Broad as it is, I'm not sure the evidence presented speaks in favour of that recommendation. An alternative view might be that the education system simply needs to be expanded massively as soon as possible, not gradually, because it is failing to provide at all for large numbers of urban students. But in the absence of more concrete evidence on urban inequalities and access, it is difficult to say precisely what kind of reform is needed. Ultimately the set of policy implications - decentralise, enhance teacher performance, strengthen local and school governance - are boilerplate and lacking in specifics, and especially in specifics that have an (urban, Tanzanian) evidence base to support them; but that is perhaps inevitable for a report of this kind. It is still a very valuable and interesting report, hopefully marking that the State of the World's Children report is just the beginning of an interest in urban deprivation at UNICEF, and that the constant "urgent" calls for better data and research will, at some point, be followed up.

New paper: children's experiences of school in Dhaka, Bangladesh

Here is my working paper (pdf) for Innocenti on children's experiences of school in Dhaka. I interviewed teenagers from middle-class families and from slum areas to try and understand whether there were processes within the school that disadvantaged children from poorer backgrounds (in addition to processes outside the school). However, I couldn't find many children who were from radically different backgrounds yet in the same school. In other words, sorting between schools - and at the secondary level, this mostly means different types of private school - is so strong that it's hard to determine whether what happens /within/ schools also reproduces educational inequality. And there were some surprising similarities across different backgrounds; for instance, the experience of violence within and outside school was quite common. Nevertheless, it was clear that the middle class teenagers interviewed here had a lot more options when things went wrong and a sense of inevitability of reaching college; while those from slums were in a much more precarious position.

Parents from both types of background were paying substantial amounts both for private school fees, even while they often questioned the quality, and for private tuition, especially in the run-up to important exams. In many ways the dreams of private-schools-for-the-poor advocates have already come true in Dhaka (at the secondary level, anyway), and it doesn't work out particularly well for anyone (except possibly the richest, who were not included in this study).

Comments welcome.

A new paper in Energy Policy (not a journal I read often) argues that people living in five slums in Gujurat, India, have a kind of hierarchy of needs. They give high priority to basic infrastructure needs until those are fulfilled then shift their aspirations towards education, healthcare and jobs. The authors have a sort-of experiment in that three of the slums had been upgraded in recent years with water, sanitation, electricity, roads and flood protection, through the Slum Networking Project.

In non-upgraded slums people's highest priorities were, indeed, water, sanitation, roads, electricity, rain protection and housing. In the upgraded ones people gave highest priority to employment, housing (again), land, health and education.

I wasn't entirely convinced by the paper's psychological theory or application of it. Priorities were measured in a relative way (by ranking). Although the stark difference in priorities between the two sets of slums is interesting to see, it's not clear that education was any more important in the upgraded slums in any absolute sense. It may be that respondents simply prioritise whatever they are most lacking. (The paper doesn't present any data on school enrolments and its regression analysis, with relative rankings in both the dependent and independent variables, doesn't seem valid.)

However the authors refer to other research suggesting that energy provision is good for health and education. Perhaps most convincing is their argument that while governments have traditionally focused on improved housing (in practice this often means forced evictions) as a "panacea to social problems," access to services, and most of all in this case to clean water, is a more pressing need as perceived by the people who live in slums themselves.

Abstracting out and accepting that the paper's broad argument probably holds some water, one problem for education is that people sometimes choose to live in places where there are few schools. The good news is that investing in infrastructure may increase school enrolments among the urban poor (e.g. because they have electric light for homework). The bad news is that these arguments might be used as an excuse by governments or agencies for underfunding education in urban areas. I hope not. A question that needs to be asked before relying on infrastructure to boost enrolment rates, is whether infrastructure is the main obstacle to accessing (not just 'prioritising' or 'aspiring to') education, rather than, for instance, a simple lack of schools, unaffordable fees, or unacceptably low quality.

"If in doubt, count"

A series of articles in Environment & Urbanization highlights the lack of data about informal settlements and how that leads to their exclusion from government services and public investments, including education. The papers focus on how the inhabitants of informal settlements can document themselves, in a series of initiatives in different countries undertaken by federations belonging to Shack/Slum Dwellers International.

In Accra, for instance, people living in an informal settlement were concerned that they were being targeted by a government resettlement programme that would shift them to the outskirts of the city, far from their places of work and with inadequate provision of education or transport. They enumerated their own settlement first in 2004, finding a much higher population than expected: around 24,000, with children under 15 making up more than 20% of the population. The enumerations gradually gained public recognition of the slum, and was eventually recognized as official data by local government. The data was used in public debates and helped to stay the threat of forced eviction, and following a second enumeration in 2006/7, it helped shift government policy away from forced evictions.

This is a continued struggle, though, as the article shows: after a change of government in 2008 the policy was about to be shifted back. Once again, community members had to campaign for evictions to be stalled so that another enumeration could be carried out. They argued successfully that they could not be relocated without first assessing their needs and numbers.

The other articles are on India, Uganda, Cape Town (South Africa), Dar es Salaam (Tanzania), Namibia, Zimbabwe and Xilapa (Mexico). There are also examples from Sri Lanka, Nepal, Fiji and Thailand. What surprises me is that none of the articles specifically highlight documentation of the shortage of government schools in and around informal settlements, and none of the projects seem to have been used to advocate for better education provision. Perhaps because it is often not within the power of local governments to open new schools? Or is education just not the top priority of these communities when faced with the more pressing threat of eviction and forced relocation?

Utilising the skills and qualifications of urban refugees

Some key findings from the Women’s Refugee Commission Report, No Place to Go But Up – Urban Refugees in Johannesburg, South Africa, show that the existing skills amongst urban refugees in Johannesburg could be utilised to potentially lift them out of poverty.

‘Forced migrants, compared to South Africans, are more likely to be vulnerable to poverty and violence […] Education levels and skill sets vary among the different forced migrant groups. Many Zimbabwean migrants have good English language skills and education, but most are found in unskilled labor: service industries, construction, painting, welding and carpentry. Forced migrants from the Democratic Republic of Congo (DRC) appear to be predominately young, urban, male and educated. Many, even those with specialized skills, work in unskilled or low-skilled work: street vending, braiding hair or washing or guarding cars. Somali asylum seekers and forced migrants are predominately young and urban; most have completed primary or secondary school, and several own businesses.’

You can download the full report here.

What do you think?

If existing skills of urban refugees are not being utilised, can increased access to education in such settings provide a path out of poverty?

Bibliography

This Mendeley group is a place where I will be keeping education and urban poverty references and anyone else can also add them there. Alternatively you can download the references as an RIS file.