Pre-pandemic, the world enjoyed the lowest prevalence of extreme poverty ever recorded…
The World Bank’s recent data showing the proportion of the global population living in extreme poverty declined from 36% in 1990 to 9.2% in 2017, should we believe this to be true?
The World Bank (WB) understands poverty as a household suffering from economic deprivation (World Bank n.d.). This measurement of global poverty by income/expenditure has a contentious past (Edward & Sumner 2014, p. 67). We suggest that the WB global poverty estimates, while flawed are not useless (Alkire 2017, p. 1). This paper poses two key areas of improvement for future estimations: (i) improving the method of analysis and the raw data, and (ii) the incorporation of alternative approaches, specifically non-monetary indicators, to broaden the scope of an inherently monetary poverty perspective. We argue that the WB’s poverty estimate may not be accurate. However, through incorporating the suggested recommendations, a richer analysis of global poverty is likely.
The WB’s first iteration of poverty estimates was founded on the works of Ravallion, Datt and van de Walle (1991). They produced an average income poverty line using household survey data from eight low-income countries. Accompanying this international poverty line was also an a-dollar-a-day measurement of $1.02 per capita grounded in the purchasing power parity (PPP) metric, adjusted according to international dollars and utilising PPP exchange rates for the year 1985 (Klasen 2018). As PPP reflects relative prices which fluctuate over time, the dollar-a-day estimate is repeated every 5-10 years in order to produce a new PPP that takes into account conversion and inflation factors (Klasen 2018). In that sense, at the time, the WB made the claim that the world enjoyed the lowest prevalence of poverty ever recorded at just 9.2% in 2017, this was equivalent to 689 million people living on less than $1.90 per day, at higher poverty margins, 24.1% of people lived on $3.20 per day and 43.6% individuals living on less than $5.50 per day (The World Bank 2020). The WB methodology for calculating the global poverty estimate has evolved over time. However, a variety of problems still beset the process.
Methodology and Data Improvement
“Missing” Statistics
The use of household surveys taken from individual countries is central to producing global poverty estimates (Alkire et al. 2017, p. 48). The World Development Report in 1990 calculated its global poverty estimate based on single household surveys for 22 countries. In 2015, the WB had more than a thousand surveys to access (World Bank 2020, p. 4). Although vast improvements have been made to both the coverage and access to data, issues are still evident. Measurement of global poverty should include everyone globally. Alkire et al. (2017, p. 48) identify that there are four key ways in which the WB’s PovcalNet statistics may exclude people: (i) PovcalNet may not cover a specific region, (ii) PovcalNet may cover a specific region, but it may not produce household survey data on household consumption, (iii) PovcalNet household survey data may not provide national coverage, and (iv) stateless, transitional peoples may not be covered. This raises issues of both underrepresentation and noncoverage. We propose the WB engages in investigations to uncover the extent to which people are “missing” from global poverty counts and, from this, find additional means of data collection to incorporate these populations (Alkire et al. 2017, p. 48).
Standardising Household Data
Household surveys are not standardised across countries, and therefore different countries utilise these surveys for different purposes (Alkire et al. 2017, p. 37). The need to improve reliability and consistency could be addressed by the establishment of a WB Joint Statistical Working Group. Whereby the group sets guidelines for the measurement of household consumption, as well as monitoring the production of data from countries regarding their household surveys (Alkire et al. 2017, p. 37). Global poverty estimates would benefit from efforts to improve data collection practices within countries and to compile individual records from surveys. These improvements would notably increase the confidence in estimates of global poverty (Dhongde & Minoiu 2011, p. 24)
Sources of Error
The process of gathering global quantitative data leaves space for multiple potential sources of error (Atkinson 2019, p. 10). The reliability of data estimates is not fully known, and therefore the WB should quote a standard error, as common practice in most national data collections. Sumner and Edward (2014, p. 79) also pose the idea that when providing estimates and forecasts of global poverty, this should be presented as a range in order to account for errors introduced by the data collection process.
Alternative Poverty Indicators
The WB as an international institution central to monitoring monetary poverty, has primarily been concerned with economic growth and therefore centres its understanding of poverty in terms of economic resources. More specifically, the institution utilises the welfare indicator ‘consumption per capita’ for its analysis of poverty (World Bank 2020). Undoubtedly, the complexities of poverty and social exclusion cannot be sufficiently determined through a single indicator relating to disposable income (Watson et al. 2017, p. 369). There are various dimensions to poverty outside of money, stressing a multidimensional approach to its evaluation.
Pham and Mukhopadhaya (2021) utilise multiple dimensions, including education, health, housing, basic services, durable assets, and economic status, in addition to income. In order to estimate poverty and vulnerability in Vietnam. They find that urgent policy direction beyond a monetary centred approach is essential for poverty identification, predication and alleviation. Moreover, Stavaren et al. (2013, p. 8) highlight the role informal institutions play in affecting well-being, poverty and economic growth. They pose a variety of social development indicators such as civic activism, gender equality, and inclusion of minorities as measures that contribute to understanding the complexity of multidimensional poverty. The failure to acknowledge poverties multidimensional element may not uncover certain populations living in poverty. Exemplified in Senegal, in which monetary poverty rates are 48.5%, compared to multidimensional poverty rates of 60%. This suggests that a multidimensional WB estimation of poverty may prove to demonstrate a far greater poverty figure (Ki, Faye & Faye 2005).
Current methods to estimate global poverty rates are flawed. We have expressed a number of criticisms of the WB’s approach, namely surrounding data collection, analysis and its monetary centered framework. We suggest given these methodological weaknesses, the claim that ‘those living in extreme poverty since 1990 has declined from 36% to 9.2% in 2017’, may very well not be the full picture. In turn, this paper has proposed recommendations for improving global poverty estimations to produce a more accurate and thorough understanding of the multidimensional poverty landscape.
References
Ahluwalia, M, Carter, N & Chenery, H 1979, ‘Growth and poverty in developing countries’, Journal of Development Economics, vol. 6, no. 3, pp. 299-341, viewed 20 March, 2021, https://doi.org/10.1016/0304-3878(79)90020-8
Atkinson, T, Alkire, S, Allen R, Bhorat, H, Bourguignon, F, Brandolini, A, Chandy, L, Chen, S, Deaton, A, Dercon, S, Ferreira, F, Foster, J, Ibáñez, Kanbur, R, Lanjouw, P, Lustig, N, Marlier, E, Mwaabu, G, Ravallion, M, Ravenga, A, Saunders, P, Sen, A, Subramanian S, & Ling, Z, 2017, Monitoring Global Poverty: Report of the Commission on Global Poverty, International Bank for Reconstruction and Development / The World Bank, WB, Washington DC.
Atkinson, A 2019, Measuring Poverty Around the World, Princeton University Press, New Jersey, USA.
Dhongde, S & Minoiu, C 2011, ‘Global Poverty Estimates: A Sensistivity Analysis’, International Monetary Fund Working Paper, IMF Institute, Washington, DC. file:///Users/tayla/Downloads/Global_Poverty_Estimates_A_Sensitivity_Analysis.pdf
Edward, P & Sumner, A 2014, ‘Estimating the Scale and Geogrpahy of Global Poverty Now and In the Future: How much difference do method and assumptions make?’, World Development, vol. 58, pp. 67-82, viewed 15 March, 2021, https://doi.org/10.1016/j.worlddev.2013.12.009
Klasen, S 2018, Measuring Poverty, Edward Elgar Publishing, Cheltenham, UK, viewed 15 March, 2021, https://www.e-elgar.com/shop/gbp/measuring-poverty-9781786436948.html
Ki, J, Faye, S & Faye, B 2005, ‘Poverty in Senegal: A Non-Monetary Basic Needs Approach’, PMMA Working Paper No. 2005-05, https://papers.ssrn.com/Sol3/papers.cfm?abstract_id=985928
Pham, A, Mukhopadhaya, P & Vu, H 2021, ‘Estimating poverty and vulnerability to monetary and non-monetary poverty: the case of Vietnam’, Empirical Economics, viewed 20 March, 2021, DOI: https://doi.org/10.1007/s00181-020-01991-4
Ravallion, M, Datt, G & van de Walle, D 1991, ‘Quantifying absolute poverty in the developing world’, Review of Income and Wealth, vol. 37, no. 4, pp. 345-361, viewed 20 March, 2021, https://doi.org/10.1111/j.1475-4991.1991.tb00378.x
Stavaren, I, Webbink, E, Haan, A & Foa, R 2013, ‘The Last Mile in Analysing Wellbeing and Poverty: Indices of Social Development’, Forum for Social Economics, vol. 43, no. 1, pp. 8-26, viewed 20 March, 2021, https://doi.org/10.1080/07360932.2013.780980
The World Bank 2020, Measuring Poverty, The World Bank, viewed 28 March 2021, https://www.worldbank.org/en/topic/measuringpoverty#2
Watson, D, Whelan, C, Maître, B & Williams, J 2017, ‘Non-monetary Indicators and Multiple Dimensions: The ESRI Approach to Poverty Measurement’, The Economic and Social Review, vol. 48, no. 4, pp. 369-392, viewed 20 March, 2021, https://www.esr.ie/article/view/819