Much of the world is deprived of poverty data. Let’s fix this
This page in:
The availability of poverty data has increased over the last 20 years but large gaps remain
About half the countries we studied in our recent paper, Data Deprivation, Another Deprivation to End are deprived of adequate data on poverty. This is a huge problem because the poor, who often lack political representation and agency, will remain invisible unless objective and properly sampled surveys reveal where they are, and how they’re faring. The lack of data on human and social development should be seen as a form of deprivation, and along with poverty, data deprivation should be eradicated.
Countries should measure poverty regularly, and have 2 or more poverty data points for every ten year period by 2030 – the target year for ending extreme poverty.
There is a growing chorus against data deprivation. Mo Ibrahim bemoans the ‘extreme poverty of data’ while Lawrence Chandy and Homi Kharas are underwhelmed and frustrated by the frequency of the global poverty estimates produced by the World Bank. Aleem Walji notes that recent poverty estimates are available for only half of the 155 countries the World Bank monitors poverty in, and a recent blog by one of us touches on this.
Finally, the recent United Nations IEAG Report “A World That Counts” argues that a lack of data can lead to a “denial of basic rights.”
Analyzing poverty data points using the World Development Indicators
We analyzed the availability of poverty data using the World Development Indicators (WDI) database, and have introduced a metric to measure the magnitude of “Data Deprivation.” The WDI database identifies 1,101 different poverty data points based on household consumption surveys across developing countries between 1976 and 2013, for which poverty measures were computed at national or international poverty lines. While more data is becoming available over time, the extent of data deprivation is considerable.
To grasp whether poverty in a country is rising or declining, at least two data points within a ‘reasonable’ time interval are required. Inspired by the IMF's General Data Dissemination System (GDDS) that calls for poverty updates every 3 to 5 years, we measure the availability and frequency of poverty data across countries in 10-year intervals. In order to meet the recommendation by the GDDS, there must be at least 2 poverty estimates in a 10-year time period.
77 out of 155 countries measured do not have adequate poverty data
The good news is that this metric of data availability points to clear improvements over time. As Figure 1 shows, the total number of countries with zero poverty data points declined from 50 during 1990-99 to 29 during 2002-11. A total of 63 out of the 155 countries surveyed have satisfactory data availability with 3 or more data points during 2002- 2011, up from 41 during 1990-1999.
Figure 1: Availability of Poverty Data in Ten Year Periods
The challenges, however, are quite clear. A total of 57 countries from the study have either zero or 1 poverty data point between 2002 and 2011. While 35 countries have 2 data points, 20 of them experienced a stretch of more than 5 years without poverty data between 2002 and 2011. In summary, about half of the countries surveyed – 77 of 155 – are deprived of adequate data, 57 of them quite acutely. Using the terms proposed in our paper: 20 countries are vulnerable to data deprivation, while another 57 have either moderate (28) or extreme data deprivation (29).
A lack of survey data precludes country-level poverty monitoring efforts and also impedes global monitoring by reducing the precision of global estimates and relying on projections.
So how do we set targets to end data deprivation?
A proposal on indicators and targets for ending data deprivation
Based on the two principles of minimum requirements for data needs and satisfactory levels of data needs for poverty, we suggest two indicators of data availability:
- Indicator 1: The number of countries with two or more data points in the last ten years; and
- Indicator 2: The number of countries with three or more data points in the last ten years.
Projecting the recent trends of the two indicators into 2030 (Figure 2), we propose the following targets:
- Target 1: End “Data Deprivation” by 2030. All countries will have two or more poverty data points in the last ten years by 2030.
- Target 2: 86 countries will have three or more poverty data points in ten years by 2030.
Figure 2: The pace of growth in Indicator 1 and 2
The implication of Target 1 is that the goal of ending extreme poverty by 2030 must be accompanied by the end of data deprivation. These targets are ambitious but possible. The proposed targets also align with the call for a ‘data revolution’ – exploiting advances in knowledge and technology, utilizing resources for capacity development, and improving coordination of efforts among key actors to mobilize sustainable development (IEAG 2014).
It is true that data can not be made available because majority of the globally available financiers do not support collection of such data! They in many cases work with Governments, and many governments are not reasy to reveal data which show the truth! Some countries have even reached a stage of enacting laws which ensure that only the state owned institutions can release data!! Therefore, deprivation will continue as long as the system of data collection, types of data to be collected are all controlled by states.!!
Thank you for your comments Dr. Tiba. Data access is an important issue to focus on in the coming years. The World Bank has been very committed to promoting this issue, e.g., through its Open Data Initiative. It also has initiatives to help developing countries start open data initiatives and to increase the use of open data in developing countries.
There is no proper definition of poverty. According to World Bank (WB)those who get an income of less $1.25 a day are considered poor. This cannot be correct. In many countries even $ 2.00 per day per person is not adequate to maintain one self. It is difficult to understand how the WB decided the value $ 1.25. This value depends on the country and sometimes on the state/region of the country. For example in Sri Lanka those who are in rural areas may manage with $1.25 per day but those in urban areas need much more per day.
Dear Dr. Weeraratna, Kindly look at the World Bank’s Povcalnet webpage (http://iresearch.worldbank.org/PovcalNet/), especially the ‘methodology’ page. This should clarify some of the concerns you have regarding the $1.25 line. Also, the page acknowledges that multiple poverty lines should be used to test the robustness of global poverty comparisons. Moreover, the page encourages users to choose their own poverty line to calculate associated poverty rates and to examine the sensitivity of poverty measures.
If I recall, the international poverty line came from averaging the national poverty lines of the 15 poorest countries in the world.
According to World Bank (WB) those who are getting a daily income less than US$ 1.25 are considered poor. This is not correct because US 1.25 per day is inadequate for a person to exist in many countries. Even in the same country the cost of living is more in urban areas than in rural areas.
There are two layers of poverty data deprivation, which the recent IAEG report discusses. The top layer, highlighted in this blog, is the absence of high level poverty indicators (both global and national), such as the poverty headcount ratio or the the poverty gap. The WDI database contains these indicators when and wherever reliable estimates are available. The base layer of deprivation is the lack of access to the underlying survey data from which the high level indicators are computed. This gap is far greater impediment to the understanding of the causes and consequences of poverty. Even when World Bank staff have access to the unit record data for the purpose of computing poverty rates or other kinds of analysis, they are often restricted from sharing the data with others who could productively use the data. And, as Dr. Tiba points out, governments frequently deny their own citizens access to the data collected in their own country.
To end data deprivation there needs to be ongoing investment in the production of high quality data -- well designed surveys conducted at regular intervals. -- and guarantees of open access with appropriate safeguards to ensure privacy and prevent improper use.
Eric Swanson
www.opendatawatch.com
Thanks Eric for your thoughtful comment and agree with you that along with data production data access is key.
The author hit the nail in the head. Moreover, while the open data movement (and even big data and data mining movements) is focused more on the most empowered, the open data movement is not much focused to the less empowered (i.e. they do not enjoy ICTs and live in rural areas). Symptoms of this are: i) The official WB Open Data site shares already aggregated and quantified surveys, which have hidden the realities in poorer neighborhoods and rural areas. While they share some microdata, such is very limited and they are nowhere machine readable, thus inhibiting processing. ii) There's no concerted effort to gather data from all NGOs and other institutions that gather data as part of their baselines; by personal experience I can tell that poor areas are surveyed by various organizations but such data is not made available nor shared with the community. iii) Why is data necessary? to satisfy paper hungry researchers? justify international development inefficient interventions? say that we have a data portal? NONE of the above. Until that data is shared with the poorest of the poor in appropriate formats so they can be empowered, we are just benefitting hidden agendas. A recent paper published in the International Political Science Review ("Evaluating the role availability") deals with the issue of empowerment and data availability. Overall, the openning of data is not meeting expectations, particularly in regards to the empowerment of those who need the most. iv) Microdata must be made available through open data portals; in fact the whole precept of the open data movement is to have access to the raw data, as it was collected from the original source. Most organizations are providing already interpreted data whose value chain from the raw data is not transparent; thus nothing can be verified. v) NSOs (or national statistical offices) are not in the open data equation. Until we grab the bull by the horns, we are just reinventing the wheel. Thus, open data movements should focus on NSOs and influencing policies that favor the sharing of microdata w/o compromising personal privacy (naturally).