Poverty is a dilemma faced across the globe by the civilization of man. Poverty can be described as living conditions that are detrimental to health, economic development, and comfort. There are different forms of poverty and different levels of poverty. Poverty is extremely dangerous because it can lead to higher infant mortality, shorter life spans, lower literacy, environmental degradation, and loss of biodiversity. In order for greater human development and overall global well-being, poverty must be eradicated. There are already efforts currently targeting poverty, including anti poverty agencies. So, how can the Big Data be used to eradicate poverty?
Amartya Sen’s definition of human development is that human development is the expansion of citizens capabilities. Sen states that freedom means increasing citizens access and opportunities to the things they have reason to value. This article relates to Amartya Sen’s definition of human development because it talks about the wealth and poverty which are usually measurements of household income and consumption. Household income and consumption are factors in the expansion of citizen capabilities. Many households around the world do not have a large enough income to meet their own needs. Due to this, measurements of poverty around the world are recorded, in order to provide data for anti poverty programs.
In the article, it states that researchers use satellite imagery and mobile data to measure and combat poverty. Examples of this are nightlight-based and daylight-based satellite measures, internet and social media activity through mobile phones, and data mining. These measures are cheaper than the use of traditional methods of data collection such as nationally representative surveys. The methods used by researchers include satellite photographs used at night, mining through social media sites and internet searches, and using digital footprints left by mobile phones.
These methods give researchers a variety of data that can be used to record poverty around the world, especially in places that are hard to reach such as third-world countries or closed off countries such as North Korea. Nightlight data is used to capture light emitted from the Earth’s surface at night. This method became available as early as the 1970s, and it used to show where the wealthy regions around the world are because they tend to shine the brightest. Studies have shown that there is a strong correlation between night light luminosity and economic productivity and growth. Another method is the mining of social media and the internet to collect data on different regions around the world. Measuring regional mobile phone use is another method to track wealth and poverty.
Although many of these methods are useful in measuring wealth and poverty, the data does not cover the whole world. An example of this is in the nightlight data collection, most extremely poor areas do not have lights at night, so data can only be collected on mostly wealthy and already developed countries. Another example of this is the collection of data through technology such as the internet and mobile phone use. Many remote and developing areas do not have the infrastructure for people to use the internet and mobile phones. People of extreme poverty do not use the internet or mobile data.
Although there are some downfalls to these nontraditional methods, they do help data collection around the world for researchers. Blumenstock’s scientific question through the article is “How might these results change the way that we measure and target poverty?”. The results are from the different data collection methods that use satellite and mobile phone data.
This paper shows the research on the measurement of poverty and inequality in Southeast Asia for the past five decades. It combines the work done by independent researchers and international agencies. The paper argues that traditional household surveys done by national statistical agencies across Southeast Asia are flawed. It states that the results for individual countries and country comparisons have been mislead due to the flaws of these surveys. It also states that the estimates of poverty and inequality are usually politically sensitive, and targets governments to create more accurate estimations through their support of the debate on measurement problems.
This paper relates to Amartya Sen’s definition of human development because of the data the paper describes. The data relates to poverty and inequality which are restricting factors in citizen capabilities. Poverty restricts citizens from many opportunities due to the affordability of these opportunities. Inequality also restricts citizens because many people are not given the same opportunities as others and as a result they are not able to be capable and active citizens.
Some sustainable development goals that are targeted in the article are good health and well-being, gender equality, reduced inequalities, no poverty, and many more. The paper wants to target many of these goals because of the poverty and inequalities faced by the people of Southeast Asia.
The article criticizes the use of household survey data because of its flawed results. It states that these surveys underestimate the incomes and expenditures of both upper and lower income groups. It does not include homeless, internally displaced, and migrant populations which are groups that are likely to be poorer than average people. On the other hand, many higher income groups may under report their income and expenditure because of the fear of tax demands. The World Bank argued that the under reporting of households produce lower measures of inequality which results in flawed data.
The paper compares many different countries in Southeast Asia across different periods of time. This includes the Philippines, Vietnam, Indonesia, Malaysia, and etc. Through this information, a more detailed view of poverty and inequality around Southeast Asia is shown. It shows that in many different countries poverty and inequality are somewhat similar and different at the same time.
The paper raises many questions about the problems of past data collection on poverty and inequality. It investigates these problems and questions the authenticity and accuracy of the data. It asks questions such as “To what extent do disparities between two sets of data reflect problems in the surveys, including poor design?” and “ Is it true that there is growing non-compliance and under reporting, not just by the very rich but across the whole spectrum of households?”. It criticizes the past data collection and estimation of poverty across many countries in Southeast Asia, and it emphasizes the importance of accurate estimation of poverty across Southeast Asia in order to combat the major problem.
This paper also combats the problem of global poverty and its measurement. It uses different materials and methods of data collection methods to map global poverty. These methods include LandScan 2004, Nighttime lights, and the poverty index and calibration. Through these methods, the group of researchers are able to estimate the number of individuals living in poverty around the world. It also demonstrates a new way of mapping poverty that can be improved over time through the calibration of poverty estimates and improvement in the accuracy of data collection methods.
Poverty is described as a chronic dilemma that civilization faces today. The paper defines poverty as “living conditions that are detrimental to health, comfort, and economic development. This relates to Amartya Sen’s definition of human development because poverty restricts citizen capabilities due to the risks in health and economic development. These risks include higher infant mortality, shorter lives, and lower literacy across the population.
The paper takes data from the World Bank and uses different methods to collect data. This data is then used to create poverty maps which are important tools for providing aid and resources. Poverty maps use different collections of data to accurately map poverty. Satellite sensors are a major source of poverty data collection, and they are extremely useful to create poverty maps.
LandScan, Nighttime lights, and the poverty index are used to collect data. LandScan is used to perform spatial allocation of census reported population numbers from models created from spatially disaggregated data. The first LandScan product used DMSP nighttime lights for the mapping of human satellites. Nighttime lights are used in the collection of data because they are useful in mapping developed or developing regions. DMSP is the US Air Force Defense Meteorological Satellite Program which uses the Operational Linescan System for mapping artificial lighting at the earth’s surface. Many studies have shown the relationship between the DMSP nighttime lights and wealth. The nighttime lights are used to build open studies of an original poverty map. Finally, the last method is the poverty index and valibration. The poverty index is calculated by dividing LandScan 2004 population count by average visible band digital number from the lights. It is able to show places of high poverty through areas with high LandScan population count and dim to no lights from the nighttime lights data.
The data collected from these methods were then used to get a percent estimate of poverty in a grid cell, and then the percent estimate was multiplied by the LandScan population grid to create an estimate of the poverty count. Through the research and data collection, the researchers were able to create a more accurate poverty map.
This paper explains the importance of data collection of poverty for improved diagnosis and policy planning to reach the Sustainable Development Goals. It recognizes the promise shown through Big Data sources such as data records and satellite imagery. The main topic of the paper is to combine disparate data sources such as environmental and mobile data to create more accurate predictions of poverty.
This paper recognizes the new strategies used to collect data, but it states the most reliable way to estimate poverty is through intensive socio economic household surveys. However, it states that this approach is costly and time consuming, and it can only cover a small sample. It introduces the growing interest of the use of Big Data to understand human development in Africa. It states that because poverty is complex, it must use multiple diverse datasets to create more accurate maps.
It analyzes the use of different data, such as satellite and GIS data, in understanding poverty. The study specifically focuses on Senegal and the use of mobile data in the form of call data records, also known as CDRs. It also uses data related to food security, economic activity, and environmental data. The objective of the researchers was to create a computational framework that integrates disparate data sources to predict Global MPI - the percentage of people who are MPI poort multiplied by the average intensity of MPI poverty across the poor - accurately. Through this an estimate for poverty was able to be created.
The materials and methods used to collect and create data were CDRs, Environmental features, and census. CDRs were analyzed to quantify the mobile use pattern of a subscriber. The environmental features were three broad categories, and these were food security, economic activity, and access to services. These three were chosen because they seem to cover most of the features that are significantly related to poverty in the literature. This data was available in raster grid or in vector format. All the data was converted to raster data and were combined to show specific commune values. Census data was used to create commune poverty statistics, and then all the data was used to create models in order to predict and map poverty more accurately.
The paper first starts out by describing poverty and its conflicts. It states that eradicating poverty is the major challenge and first target of the Sustainable Development Goals. In order to eradicate poverty, we must first understand the causes and improve are resources to combat it. One way to measure poverty is through examining indicators of living standards for a population, but each approach to calculating these indicators have advantages and disadvantages. There are a variety of indicators and each one provides different characteristics of the population. As a result, the researchers take different types of data and overlap them to evaluate and estimate three different measures of poverty.
The materials and methods used to estimate measures of poverty were spatial scale and data processing, poverty data, CDR and RS data, Covariate selection, and prediction mapping. Each material or method has their strengths and weaknesses, and we can see that through the paper. Spatial scale and data processing were used to look at mobile phone data within urban areas. Poverty data was analyzed through the use of datasets representing asset, consumption, and income-based measures. CDR and RS data were analyzed through mobile phone metadata collected from November 2013 and March 2014. Covariate selection data was used to create models that were then used in three poverty measures for national, urban, and rural strata. These measures were then used to explore differences in factors related to urban rural poverty. The last method used was prediction mapping, and this used the models from covariate selection data to predict the three poverty metrics at unsampled locations across the population.
Through the use of the results of the materials, methods, and models, a first attempt was made to build predictive maps of poverty using combinations of CDR and RS data. The findings also support the strong relationship between socio-economic measures and night-time light intensity. The results were also used to produce accurate, high-resolution poverty maps. This research introduces the potential for a stronger spatial distribution of poverty measurement and the foundation for strategies to eradicate poverty. It is an introduction to the future of the use of Big Data and its role in combating poverty globally.