Conceptual and methodological issues in defining and measuring absolute poverty

Saraceno, Chiara Article — Published Version Conceptual and methodological issues in defining and measuring absolute poverty Rivista Italiana di Economia Demografia e Statistica Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Saraceno, Chiara (2024) : Conceptual and methodological issues in defining and measuring absolute poverty, Rivista Italiana di Economia Demografia e Statistica, ISSN 0035-6832, Società Italiana di Economia Demografia e Statistica, Firenze, Vol. 78, Iss. 4, pp. 9-20, https://doi.org/10.71014/sieds.v78i4.405 This Version is available at: https://hdl.handle.net/10419/315541 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0 Rivista Italiana di Economia Demografia e Statistica Volume LXXVIII n.4 Ottobre-Dicembre 2024 CONCEPTUAL AND METHODOLOGICAL ISSUES IN DEFINING AND MEASURING ABSOLUTE POVERTY Chiara Saraceno Abstract. The notion of absolute poverty recalls that of basic needs, which, however, is far from having a consensual meaning. The author first discusses two main contrasting definitions of basic needs - one based on an idea of pure subsistence, the other based on human rights and the capability approach. The author then outlines the conceptual and methodological issues involved in intraand cross country as well as over time comparison. In the final paragraph, the author suggests that in order to overcome the limitations of any poverty measure that ignores intraand cross-country differences in access to human and social rights and in the availability of common goods, the author suggests that survey-based data on income or consumption should be integrated with data on selected functionings at the individual and household level (e.g., education, health, housing), which in turn should be contextualized based on the national and local availability and accessibility of public goods that are defined as essential for supporting those functionings. 1. Introduction Absolute, or extreme, poverty and how to measure it until recently have been considered an issue concerning exclusively developing countries. Within the developed, particularly Western countries the focus has been rather on relative poverty, although with the important exception of the United States. This choice is based on the idea that, in these countries, extreme lack of means of subsistence had been largely overcome at least since the post-World War II years, due to the combination of economic development and welfare state arrangements. Furthermore, as noticed by Blank (2008), who criticized the US approach on this ground, absolute poverty lines typically lag behind the average living conditions in rapidly growing economies, an argument that is becoming important also at least in some of the developing countries. The reference point for assessing poverty from a relative perspective, therefore, is not pure subsistence but the average level of living. Following Townsend (1962, 1979), relative poverty concerns the inability to adequately participate in the society one lives in for lack of adequate resources. The EU indicator of “at risk of poverty” (AROP) is also based on the relative poverty concept. [ https://doi.org/10.71014/sieds.v78i4.405 ] 10 Volume LXXVIII n.4 Ottobre-Dicembre 2024 In recent years, however, some dissatisfaction with this concept has emerged, for different reasons. One concerns its comparability when countries with great differences in level of living are involved. This is particularly true when relative poverty is expressed, as in the EU indicator, in terms of percentage (50% or 60%) of mean or median national income. A poverty line set at 50% or 60% of mean or median national income, in fact, does not always indicate a similar, or comparable, situation across countries and it risks representing as non-poor individuals and households in one country who would be severely poor in another. In order to overcome this paradox and adequately assess and compare levels of poverty, a common standard must be identified and its value translated in PPS, as suggested by Atkinson (1998, see also Atkinson et al. 2002). Two other reasons for the partial dissatisfaction with the concept and measurement of relative poverty concern instead its efficacy in actually detecting and measuring poverty. In addition to being very sensitive to the economic conjuncture (see e.g. Jenkins et al. 2013), it seems rather a measure of inequality than of straightforward poverty. Because of this dissatisfaction, there has been an effort to integrate the measurement of relative poverty with other indicators. At the EU level, Eurostat has developed a multidimensional indicator of at risk of poverty or social exclusion (AROPE) which, alongside relative income poverty, includes living in a low work intensity household and suffering severe material and social deprivation. This last sub-indicator may be understood as a, partial, indicator of absolute poverty in so far it is based on a number of goods and consumptions, lack of access to a number of which (7 out of 13) is defined as severe deprivation. Furthermore, with the Measuring and Monitoring Absolute Poverty (ABSPO) project, the European Commission and Eurostat have assessed not only the feasibility, but the opportunity to integrate the present indicators and ways of measuring poverty, with an absolute monetary poverty measure that represents constant or comparable purchasing power over commodities across countries and time periods (European Commission, 2021). At present, however, within the EU member countries only Italy since 1997 has an official absolute poverty measure alongside the relative one. 2. Different approaches to defining absolute poverty The notion of absolute poverty recalls that of basic needs, which, however, is far from having a consensual meaning. Different understandings and definitions of what are basic needs, and what are the goods needed to satisfy them, may result in different definitions, and measures, of absolute poverty. Within the international debate we, may find two main contrasting definitions of basic needs. The most consolidated, but also simpler, definition is based on a concept [ https://doi.org/10.71014/sieds.v78i4.405 ] Rivista Italiana di Economia Demografia e Statistica 11 of pure subsistence. It defines absolute poverty as the lack of sufficient resources to secure basic life necessities, including amongst others safe drinking water, food, or sanitation. Not surprisingly the term absolute poverty thus defined is often interchangeable with that of extreme poverty. The most extreme use of this definition may be found in the World Bank estimates of the incidence of poverty across the world, where the international extreme poverty line is based on the lines of the group of poorest countries. According to this definition, are absolutely or extremely poor those who live with less of the equivalent in PPS of 2.15 dollars a day at 2017 prices. In 2020, its value was € 1,39 a day in Italy, €1.41 in Portugal, 7.49 yuan in China, 22.49 pesos in Mexico, 355.18 naira in Nigeria. The amount is so low that even those who are a little above it may experience difficulty in surviving, not to say live decently and with dignity. It should be noted, however, that, following recommendations by Atkinson (2017), the World Bank is now also reporting estimates based on a ‘societal poverty line’ which combines absolute and relative elements (World Bank, 2018). Other exercises that share a “survival” concept of absolute poverty are the International Food poverty line proposed by Kakwani and Son (2006), the Minimum Income for Healthy Living proposed by Morris and others (2000), the international poverty measurement proposed by Allen (2017) as conceptually and methodological superior to that of the World Bank. All these attempts struggle with the issue of what is needed to “survive” in a given time and society, and not merely live “hand to mouth”. In this perspective, it is worthwhile remembering that the very initiator of the absolute poverty approach, Rowntree (1937), in constructing the basket of essential goods in 1930s UK, included also tea and tobacco, which did not have any nutritional and healthy-wise value, but were psychologically and socially important for English men at the time. Rowntree forcefully argued that “working people are just as human as those with more money. They cannot live ‘on a fodder basis’. They crave for relaxation and recreation just as the rest of us. But … they can only get these things by going short of something which is essential to physical fitness, and so they go short.” (Rowntree 1937, pp. 126-127). This opinion seems to be shared by all attempts to define absolute poverty lines in the developed countries, in so far, as pointed out also in the final ABSPO Report, they refer to benchmarks well above pure subsistence and include the accessibility to consumptions that are considered not only desirable, but necessary in a given country. The second main definition of absolute poverty is based on the Human rights and capabilities approaches (although there is some difference between the two). It therefore conceives absolute poverty as the inability to reach minimum acceptable levels in the fundamental rights, or functionings (Sen, 1992; Tiraferri, 2008). These, of course, include nutrition, health (including also infant mortality protection), [ https://doi.org/10.71014/sieds.v78i4.405 ] 12 Volume LXXVIII n.4 Ottobre-Dicembre 2024 shelter, but also education, dignity and above all the possibility to choose what kind of life to have; thus, they also include the degree to which human rights are granted or on the contrary infringed upon. Severe constraints in accessing education, health care, the regular labour market, lack of protection from exploitation, high inequalities in the risk of infant mortality depending on one’s position in the social stratification and/or (for instance in Italy with regard to infant mortality; see De Curtis and Simeoni, 2021) region of residence – all these dimensions are perceived as important as the economic condition in determining absolute poverty and should be taken account of when measuring it. Towards this end, Alkire and Foster (2011) have developed a general measure of multidimensional poverty based on dimensions included in the Human development Index. Their approach has been implemented in the construction of the Multidimensional Poverty Index (MPI) by the Oxford Poverty and Human Development Initiative (,) and the United Nation Development Programme for their joint Annual Report on Global Poverty (OPHI/UNPD 2023). The British New Economic Foundation (NEF) Think tank, as well, proposes a rights-based poverty line (RBPL) that is based on the estimated relationship between income and wellbeing indicators, which refer to distinct economic and social rights. The minimum threshold for each indicator is universal, but the needed income to achieve it varies across countries. Interestingly, this proposal suggests that the role of the availability of public goods in reaching and eventually overcoming the poverty threshold in a given dimension should be taken account of, although without clear methodological indications (New Economics Foundation, 2010). Atkinson and Bourguignon (1999) have proposed an integration between the two main definitions of absolute poverty and underlying basic needs, distinguishing between two levels of capabilities. The first one refers to subsistence, the second to social functionings. According to these two authors, however, the first level refers to absolute poverty, the second to relative poverty. Sen, on the contrary, argues that achieving a minimum level of social functionings is also necessary in order to have a life with dignity. The impossibility to achieve this minimum level, therefore, represents a form of absolute poverty. Relativity, according to Sen, concerns only the resources needed to adequately “function” and achieve a minimum acceptable level of capabilities, since they vary depending on the context. “Being in good health”, or “having the necessary minimum of education” are absolute needs, with the same status as, for instance, access to clean water, shelter, clothing and food. But the definition of thresholds and the actual means to achieve them depend on available hygienic, health and school infrastructures in a given context.”The characteristic feature of ‘absoluteness’ is neither constancy over time, nor invariance between different societies, nor concentration merely on food and nutrition. It is an approach of judging a person’s deprivation in absolute terms rather [ https://doi.org/10.71014/sieds.v78i4.405 ] Rivista Italiana di Economia Demografia e Statistica 13 than purely relative terms vis à vis the levels enjoyed by others in the society” (Sen 1985, p. 673). He therefore proposes to define poverty in terms of capability failure: “the failure of basic capabilities to reach certain minimally acceptable levels” (1992, p. 109). Doyal and Gough (1991) with their concept of “combined capabilities” and “intermediate needs” have further worked with this idea of universal basic needs to make it more helpful for social policy. Combined capabilities and intermediate needs do not exist as an abstract potential but are mediated by material resources. In Doyal and Gough’s words (1991, p. 157), they are “the crucial bridge between universal basic needs and socially relative satisfiers”. The object of evaluation, from the point of view of the assessment of poverty, are not needs, or capabilities, in the abstract, but the availability of and access to combined capabilities or intermediate needs (such as, for instance, adequate nutritional food and water, adequate protective housing, basic education, reproductive freedom). The United Nations in 1995 adopted a definition of absolute poverty that aims at combining a survival and a human rights-based approach, as well as an absolute and relative perspective. In this definition, absolute poverty is characterised as “severe deprivation of basic human needs, including food, safe drinking water, sanitation facilities, health, shelter, education and information”. It is related to social services as well as income. It is part of, but distinguished from, a broader notion of “overall poverty”. This latter refers to the total number of people living in poverty in a country since they lack a number of goods and/or opportunities, which range from “lack of income and productive resources sufficient to ensure sustainable livelihood” to “increased morbidity and mortality from illness”, to “social discrimination and exclusion”, “lack of participation in decision-making and in civil, social and cultural life” (UN, 1995, par. 19). It should be noted that the “father” himself of the relative poverty approach, Townsend, was involved in drafting this definition and in the ensuing campaign for a global social floor (Yeates and Deacon, 2011). The definition of minimum living standard as adequate participation in society proposed by the cited EU sponsored ABSPO project also may be considered as belonging to the human rights and capability approach. As the Report underlines, in fact, it is consistent with the widespread view of social participation as a summary indicator of both individual well-being and the fulfilment of one’s social rights (Kahneman et al., 1999; Lister, 2004, 2014), as well as with a poverty measurement that focuses on individuals’ effective freedom, agency and attainable societal roles in explaining material and social deprivation (Sen, 1985, 1987; Nussbaum, 2000). Furthermore, the concept of adequate social participation is generic and flexible enough for the needs of international measurement where countries of different socio-economic backgrounds are compared. [ https://doi.org/10.71014/sieds.v78i4.405 ] 14 Volume LXXVIII n.4 Ottobre-Dicembre 2024 The final ABSPO Report, however, admits that “The concept of adequate social participation is somewhat elusive and needs to be made operational through a series of practical measurement choices. These should clarify what social participation means, how adequacy is defined, and where the boundaries of the relevant society are drawn in each particular context” (European Commission 2021, p. 33). And it adds: “There is no agreement on these topics in the theoretical literature: many alternative concepts of social integration and participation are used simultaneously, while the study of social adequacy and societal boundaries provides little effective guidance for practical measurement” (Ibidem). 3. Context specificity: conceptual and methodological problems As Ravallion pointed out some years ago (2016), we may observe a growing compatibility between indicators of absolute and relative poverty precisely because of the awareness of the context specific quality of absolute poverty indicators. Context specificity, however, raises issues of cross-country, intra-country and cross-time comparability. There may be a more or less important cross-country variability both in the way “fundamental needs” are perceived and in the means to address them even within a group of countries belonging to the same political and institutional space. The difficult consensus on the list of “deprivations” for the AROPE indicator within the EU and the (political) impossibility to rank them testify both the different starting points and the diverse cultures characterizing these countries. Even calculating the cost of a comparable budget may be difficult, as it emerges in the interesting experiments of developing comparable “reference budgets” across the EU (e.g., Cantillon et al., 2019). In order to calculate their economic value, de facto, either average prices at the national level or the prices in one or more cities within each country are considered, ignoring within country variation, which may be significant, particularly concerning housing, but also transportation and, in the case of the rural/urban divide, even food. Comparability across time must also address the issue of variability of how “fundamental needs” are perceived and framed as well as changes in the means to address them. The higher level of literacy and schooling required over time in order to obtain a job and move competently in society, easier access to clean water, up to tap water, changes in housing and communication technologies - these and other changes may push in absolute poverty those who cannot access them, not only because they cannot reach the new minimum adequate level of living, but because these changes impact also on resources that were previously available, reducing or eliminating them. Fountains where one might wash her/his laundry disappear. Today this is happening also with public phone boxes. Rabbits and chickens can no [ https://doi.org/10.71014/sieds.v78i4.405 ] Rivista Italiana di Economia Demografia e Statistica 15 longer be raised domestically unless one lives in a rural setting. Spaces for subsistence agriculture diminish and self-building one’s own house in open “free” spaces impossible or illegal. Until the mid-1960s, in Italy, ownership of a laundry machine and refrigerator was an indicator of a good economic status. Today, lack of them may be an indicator of severe deprivation. Analogously, in the 1950s and 1960s, one might live decently without a (traditional) phone. Today a cell phone and internet access are required even for applying for social assistance when poor. It should be noted that the only developed country that uses exclusively the absolute poverty approach, for its official estimates, the US, takes account of variation over time only with reference to inflation, that is of variations in the cost of a basket of goods that has remained the same since 1960. In Italy, instead, also the “basket of essential goods” has been revised, first in 2009 and then again in 2022. The concept of absolute poverty underlying the Italian measurement method (see the methodological note in ISTAT 2023 and the contributions to this Special Issues of the Rivista Italiana di Economia Demografia e Statistica), that is how the basket of goods has been individuated, goes beyond that of subsistence for two reasons. First, it is multidimensional and not driven exclusively, as the US one, by the food component. Second, it is carefully contextualized in time and social space with regards to the goods that it considers as necessary. It may not, however, be considered as based on the capabilities or human rights approach. Expenditure for consumption, in fact, although more adequate than income to assess the level of wellbeing or deprivation, reflects only the cost of goods and services, but it ignores the variability in the availability of public and private services as well as of other important contextual dimensions (e.g., environmental quality, security, labour market opportunities) that impact on the quality of life of individuals and households (Brandolini, 2011; D’Alessio, 2018). Scarcity or absence of early child education and care services, of full-time schools, an inefficient health service, lacking and/or inefficient public transportation, lack of parks and public playgrounds, pollution, living in insecure neighborhoods – all these may substantially reduce the quality of life of the poor. Furthermore, they may differentiate the poor living under these circumstances with regard to the capacity to satisfy their basic needs not only from those who are better off, but also from individuals and households who have the same economic conditions, but live in regions, municipalities that have a better dotation of public goods and in neighborhoods that are more secure and not polluted. [ https://doi.org/10.71014/sieds.v78i4.405 ] 16 Volume LXXVIII n.4 Ottobre-Dicembre 2024 4. The unresolved issue of how to take account of the value of public goods: the need for a new approach Whether and how to include the presence or absence of public goods and their value, not only in economic, but also in wellbeing terms, raises both comparative and methodological problems, when looking at the individual and household, not societal level. It would require a detailed analysis of national and sub-national contexts on the availability, distribution, rules of access to various kinds of public goods on the one hand of whether and to what degree individuals and households have actual access to them depending on their economic and social circumstances. For instance, when first ISTAT started measuring absolute poverty, it did not include expenditure for health based on the fact that, in principle, in Italy access to health services is universal. Later, however, it decided to include also this kind of expenses not only because the National Health services introduced users’ fees, but also because of an increasing recourse to private health services. Yet, the question remains open on, first, whether, where, to what degree this recourse is dictated by lacking or inefficient public services, second, what happens when lack of public services may not be compensated with out-of-pocket expenditure because of insufficient economic means. “Health poverty” simply remains invisible. The same happens with childcare services and full day schools (which include also cafeteria services). These public goods not only are not offered on a universal basis but are also quite unevenly distributed across the national territory (ISTAT, 2020). Thus, many households and their children do not have the “choice” to use them. This scarcity, in turn, tends to strengthen the well-known phenomenon of children of dual worker, well educated, middle-high class parents disproportionally using these services, even when they are public, compared to single earner, low income and low educated parents (Pavolini and Van Lancker, 2018). Similar reasoning may be applied to the availability of other public goods, such as parks, transportation, accessible sport facilities, public libraries and so forth Even the indicators proposed by the ABSPO project do not include the availability of, and access to, public goods. Yet, as Lanau et al. (2020) observe: “Where universal provision of basic services is lacking, current approaches to poverty measurement may result in underestimates, thereby raising comparability and identification issues”. Baldini and others (2014) some years ago, suggested to take account of the role of public goods in contributing to the wellbeing of individuals and households by imputing their estimated value. But, even if this solution were theoretically adequate, which is at least controversial, it requires beforehand to know whether the considered public goods are actually accessible and under what conditions to the individuals and households to whom that value is imputed. [ https://doi.org/10.71014/sieds.v78i4.405 ]