SHARE

(C0ntinued from last week)

In this Q&A, David Newhouse a senior economist with the World Bank takes questions on how poverty is measured and why Sri Lanka needs to update its current poverty line.

Q: What is the spatial distribution of poverty and how has it shifted over time in Sri Lanka?
Data from the 2011 Census of Population and Housing (CPH) and the 2012/13 Household Income and Expenditure Survey (HIES) reveal differences in the incidence of poverty across Sri Lanka’s 25 districts and 300 odd divisional secretariat divisions.

The data reveals three main pockets of poverty. The first are the former conflict districts in Northern Province, Mullaitivu (28.8 percent), Mannar (20.1 percent), and, to a lesser extent, Kilinochchi district (12.7 percent). The second is Batticaloa district (19.4 percent) in Eastern province, and the last one is Moneragala district (20.8 percent) in Uva province.

Between 2002 and 2012/13, different districts also reported different rates of poverty reduction – the most notable results came from Kalpitiya, Mundel, and Vanathawilluwa divisions in Puttalam district, as well as in Badulla and Hambantota districts. In contrast, most DS divisions in Moneragala, Colombo, and Gampaha made little progress in reducing poverty. In fact, many pockets of high poverty existed even in affluent districts, including Colombo.

Q: The report notes that the highest poverty rates do not necessarily contain the largest number of poor people. Could you elaborate on that?
It is also important to keep in mind that although these pockets contain the highest incidence of poverty, that most of the poor live in populous and relatively affluent districts such as Kurunegala, Ratnapura, and Kandy. Kurunegala, for instance, is home to 7.7 percent of the country’s poor people even though only 6.5 percent of its population lives under the official poverty line. In contrast, Mullaitivu and Mannar, where estimated poverty rates are very high (28.8 and 20.1 percent, respectively), account for only 3.4 percent of poor people nationwide due to their small populations. The same is true at the DS division level, where large numbers of poor exist in Gampaha, Nuwara Eliya, and Ratnapura districts. This distinction is important because investment in infrastructure and other public goods may more effectively be targeted to areas where most of the poor live.
Q: What are the underlying causes of regional disparities in wealth and opportunity?
More research is needed to understand why there can be such dramatic variations even between districts that adjoin each other. What we know is that the poor and near-poor tend to be rural, young, and disconnected from productive earnings opportunities.
The rural sector accounted for over three quarters of the country’s population and over 85 percent of poor Sri Lankans nationwide in 2012/13. HIES data highlights an issue with educational qualifications among the bottom 40 percent, and unemployment in this group tends to be exceptionally high.  If this pool of abundant, young, and cheap labour is not fully utilized, the combination of high youth unemployment and a young poor population has the potential to threaten social stability.

Critically, among those poor and near-poor that are employed, a large proportion is engaged in agriculture—a field with typically fewer opportunities to add value to products and lower wages than service or industrial jobs. Agriculture still employed about 28 percent of the working population in 2015. This reinforces the case to accelerate Sri Lanka’s structural transformation, and ensure that young people have the means and ability to obtain work in more productive jobs in the industry and service sectors.

Q: What are non-monetary forms of poverty and why should they be considered complementary indicators when analyzing the welfare of various communities?
It can be difficult to get experts to agree on a single definition of poverty. One of the fundamental questions is whether household consumption is the best yardstick to measure poverty, or whether we should consider other aspects such as health, access to education and quality of housing when determining the welfare status of a household.
In Sri Lanka, one of the areas where we find it useful to measure non-monetary poverty is in the Estate sector. The poor and near-poor in the Estate sector are mainly young, Hindu Tamils, working in tea estates in the agriculture sector, and living in free housing units provided by estate owners. Monetary poverty rates in the Estate sector saw a steep decline, from 28 percent in 2006 to around 10 percent in 2009/10 and 2012/13.
Non-monetary indicators reveal cause for concern. For example, only 5.6 percent of the estate poor owned a house in 2012/13, and only 11.3 percent of the Estate near-poor do. In contrast, 90.4 percent of the rural poor owned a house. When it came to education, only 2.3 percent of adults living in the Estate sector had completed secondary school in 2012/13, as compared to 8.8 percent in the rural sector. These figures suggest that improving public services for Estate sector residents remain a priority.

Q: What can high resolution satellite imagery reveal about economic welfare? Can the use of such images help compensate for data gaps?
Despite the best efforts of national statistics offices and the international development community, local estimates of poverty and economic welfare remain infrequent. Partly, this reflects the fact that census data are collected infrequently – typically once every ten years. In some other countries, household survey data measuring economic welfare is collected once every five or ten years.

Satellite imagery has generated considerable enthusiasm as a potential supplement to household data that can help fill these severe gaps. There are two ways one can use satellite imagery for this. One is to look for ‘object features’ in the images such as the number and size of buildings, type of farmland (plantation vs. orchard), as the number of cars  the type of roofs, the prevalence of shadows, roads, and road material, along with textural measures of the type mentioned above. Texture and spectral features can also computed directly from the imagery. A second approach is to train computers to predict poverty directly, using the same technology that computers use to recognize faces or other objects.  Since direct poverty prediction is not always accurate, ongoing research is still trying to better understand the relative merits and requirements of each approach.
In our study, both object and texture features were matched to household estimates of per capita consumptions imputed into the 2011 census (the same ones used by the DCS to generate DS Division estimates of poverty), for 1,251 GN Divisions. The satellite features track poverty rates closely, as they explain around two thirds of the variation in welfare. This suggests that features extracted from high-resolution imagery could be paired with household survey data to generating more frequent poverty maps. We plan to explore opportunities to partner with the DCS to apply this new technology to update the DS Division poverty estimates when the 2016 household income and expenditure survey becomes available.

Q: How are ‘poverty maps’ used by policy-makers and various ministries?
The poverty maps of Sri Lanka have had a wide impact. Perhaps one of the most important impacts was the use of the maps by the Ministry of Samurdhi to select 113 of the poorest Divisional Secretariats (DSs) when the ministry initiated the reform of the Samurdhi transfer programme for the poor.

This did not happen in isolation; it was an outcome of the tireless efforts of the DCS to disseminate the results of the poverty mapping exercise to the public and, especially, government officials. Now, many officials are aware of poverty maps as important instruments for identifying and targeting poor areas.

The Sri Lanka poverty maps are accepted widely because the DCS has taken this lead in dissemination. DCS staff are able to explain the technical aspects of poverty maps and deal with the political sensitivity of the maps. The World Bank’s long-term commitment to capacity building within the DCS, which began well before the poverty mapping exercise, and our role in their dissemination have also contributed to this success.

Q: What can be done to build on Sri Lanka’s successes in poverty alleviation?
Some of Sri Lanka’s success can be attributed simply to rising international prices for food and tea that raised earnings in agriculture, and strong domestic aggregate demand that boosted economic growth. However, these factors cannot be relied upon to continue in the future, and it is rapid structural transformation and increased agglomeration that has the most potential to sustain poverty reduction in the future.

Efforts   to further improve living standards of the poor should focus on promoting further structural transformation and urbanization. Roughly 28 percent of the workforce, and about half of the working poor, toil in the agriculture sector. Many of the poor live in peri-urban areas – over half of the poor are estimated to live within 30 km of a main agglomeration area. Policies that help connect these workers to productive employment opportunities off the farm can contribute to sustainable poverty reduction. This is a long term agenda, however, and it will take years if not decades to reduce agricultural employment below 10 percent.

In the meantime, poor farmers and agricultural labourers need help to generate the income to allow them to invest in the human capital of their children. Some of these farmers were hit hard by bans on slash and burn agriculture. This was necessary to protect the environment and regulate land use, but in some cases limited agriculture to one growing season per year. Farmers and agricultural labourers could benefit greatly from practical infrastructure investments, such as water storage tanks, irrigation facilities, fertilizer and seeds, electricity, and roads that would make them more productive. In addition, microfinance programmes that offer loans at reasonable interest rates that can be repaid post-harvest, rather than every month, could also facilitate productive investment among smallholder farmers.

Although 95 percent of the population can now access electricity, more cost-effective power generation will be required to support further productivity growth. Effective governance of cities will also be important to enhance the ability of workers to take advantage of more productive job opportunities in and around urban areas.

Sri Lanka also has an admirable history of strong social assistance programs which are well-worth continued investment. These along with multisectoral interventions targeted to the remaining pockets of poverty, can help support the existing poor. However, we also need to think long term. Like many other countries Sri Lanka needs to prepare for the challenges of an aging population and a growing middle-class.
(World Bank)