Fact-checked by Grok 2 weeks ago

Multidimensional Poverty Index

The Multidimensional Poverty Index (MPI) is a composite that quantifies acute multidimensional poverty by assessing household deprivations in three dimensions—, and —using ten weighted indicators derived from survey data, where individuals are deemed multidimensionally poor if deprived in at least one-third of the total weighted indicators. Developed by economists Sabina Alkire and James Foster via their Alkire-Foster (AF) counting methodology, which computes both the incidence (headcount ratio) and intensity of poverty to yield an adjusted headcount measure ranging from 0 to 1, the MPI provides a flexible framework adaptable to national contexts while enabling cross-country comparisons. First introduced by the (UNDP) in its 2010 in collaboration with the Oxford Poverty and Human Development Initiative (OPHI), the global MPI has tracked acute poverty affecting over 1.1 billion people across more than 100 developing countries as of recent estimates, highlighting deprivations often overlooked by income-based metrics alone. Its adoption in policy has facilitated targeted interventions, such as in India's national MPI aligned with , though empirical critiques note methodological sensitivities including arbitrary deprivation cutoffs and equal weighting assumptions that can distort rankings, particularly inflating apparent poverty in low-income settings compared to correlation-adjusted alternatives.

History and Development

Origins and Theoretical Foundations

The theoretical foundations of the Multidimensional Index (MPI) originate from Amartya Sen's capabilities approach, articulated in works from the late 1970s onward, which reframed as a shortfall in individuals' freedoms and capabilities to achieve valued functionings rather than mere income deficiency. Sen argued that unidimensional monetary metrics, such as income lines, inadequately capture 's causal structure by neglecting deprivations in non-income domains like , , and , where empirical data from household surveys reveal persistent overlaps uncorrelated with income alone. This critique, echoed in Anand and Sen's 1997 analysis, emphasized that aggregating solely via income overlooks the multidimensional nature of human deprivation, potentially misleading policy by ignoring how deprivations compound across life aspects. Building on Sen's framework, Sabina Alkire and James Foster developed the MPI's core methodology in the mid-2000s at , motivated by the need for a practical, data-driven tool to quantify these overlapping deprivations using counting-based techniques rather than utilitarian averaging. Their seminal 2007 paper proposed an identification strategy that counts weighted deprivations across dimensions, extending axiomatic poverty measurement principles to respect the distinct, non-substitutable impacts of deprivations in areas like schooling and sanitation, as evidenced by discrepancies in cross-national datasets where monetary poor often escape non-monetary hardship and vice versa. This approach prioritized empirical robustness over income-centric proxies, addressing Sen's call for measures sensitive to the intensity and breadth of poverty experienced by individuals. The Oxford Poverty and Human Development Initiative (OPHI), co-founded by Alkire in 2007, institutionalized this work by integrating the Alkire-Foster method with Sen's capabilities paradigm, fostering research that grounded abstract theory in verifiable survey indicators to reveal poverty's true incidence beyond GDP correlations. Unlike earlier multidimensional efforts reliant on composite indices that averaged dissimilar metrics, the MPI's foundations emphasized decomposability and subgroup analysis to trace causal pathways of deprivation, enabling targeted interventions informed by direct evidence of unmet basic needs.

Launch and Initial Adoption

The Multidimensional Poverty Index (MPI) was formally launched on July 14, 2010, through a collaboration between the (UNDP) and the Oxford Poverty and Human Development Initiative (OPHI), marking its introduction in the 20th anniversary edition of the UNDP . This debut complemented the report's by providing a measure of acute multidimensional poverty, drawing on the Alkire-Foster methodology to assess deprivations in health, education, and living standards. The inaugural global MPI applied to 104 developing countries, encompassing roughly 5.2 billion people or 78% of the world's population at the time, with data sourced primarily from (DHS), (MICS), and World Health Surveys conducted between 2000 and 2008. It identified approximately 1.75 billion individuals—one-third of the covered population—as multidimensionally poor, highlighting the prevalence and intensity of overlapping deprivations across these nations. Initial adoption occurred swiftly among governments seeking tools for poverty targeting beyond monetary metrics. pioneered an official national MPI in 2010, establishing it via normative framework under the National Council for the Evaluation of Social Development Policy (CONEVAL) to guide social program allocation and evaluation. similarly adapted the approach for national use shortly thereafter, with other countries like exploring implementations to inform policy. In , the MPI gained traction through early analytical applications to 2011-12 survey data, enabling regional decompositions that supported evidence-based interventions in high-poverty areas. These early integrations underscored the index's utility for disaggregated , though national variants often customized dimensions to local contexts.

Evolution and Global Implementation

Following its launch in 2010 by the Poverty and Human Development Initiative (OPHI) in collaboration with the (UNDP), the Multidimensional Poverty Index (MPI) has undergone annual updates to refine its scope and coverage. These reports have progressively expanded data availability, drawing from household surveys such as Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Surveys (MICS), to track acute multidimensional across developing countries. By 2025, the global MPI encompassed estimates for 109 countries, representing a of approximately 6.3 billion, with consistent indicators enabling assessments of changes over time. The 2024 and 2025 global MPI reports estimated that 1.1 billion people—about 18.3% of the covered —lived in acute multidimensional , highlighting persistent challenges despite reductions in some regions. To enhance cross-country comparability and longitudinal analysis, OPHI introduced harmonized estimates in 2025, restricting computations to indicators consistently available across survey waves and countries, thereby addressing variations in and survey design that previously limited reliable trend comparisons. This methodological refinement supports more robust evaluations, though it depends on the frequency and completeness of national surveys, which remain uneven in low-income settings. Nationally, over 40 countries have adapted the MPI framework to local contexts, often customizing dimensions, indicators, and weights to reflect priorities such as housing quality or employment access. implemented one of the earliest national versions in 2011, incorporating deprivations aligned with its goals. In , the MPI draws from the 2022 Bhutan Living Standards Survey and integrates elements resonant with principles, emphasizing psychological wellbeing alongside standard health, education, and living standards indicators. These adaptations facilitate targeted interventions but require ongoing to maintain relevance. Since the adoption of the (SDGs) in 2015, the MPI has been integrated into global and poverty monitoring, particularly for SDG Target 1.2, which calls for halving the proportion of people in multidimensional by definitions. OPHI's expansions of the MPI database, leveraging MICS data from , have enabled disaggregated analyses that complement monetary metrics and inform SDG progress tracking in over 100 countries. This broader implementation underscores the MPI's role in evidence-based policymaking, contingent on sustained investments in survey infrastructure for timely, comparable data.

Methodology

Dimensions, Indicators, and Weights

The Multidimensional Poverty Index (MPI) employs three dimensions—, and —each weighted equally at one-third of the total index to reflect deprivations in basic human capabilities. These dimensions aggregate ten binary indicators, where households are deemed deprived or not based on specific cutoffs derived from international standards such as WHO nutritional guidelines and DHS survey definitions. Weights are assigned equally within dimensions: the and dimensions, with two indicators each, allocate 1/6 per indicator, while the dimension, with six indicators, allocates 1/18 per indicator, ensuring dimensional parity despite differing numbers of metrics. The indicators and their deprivation cutoffs are specified as follows:
DimensionIndicatorDeprivation CutoffWeight
HealthNutritionAny adult (19–70 years) with BMI <18.5 kg/m², or any child under 5 with height-for-age or weight-for-age z-score <-2 SD, or child (5–19 years) with BMI-for-age z-score <-2 SD.1/6
HealthChild MortalityAny child under 18 died in the household in the past 5 years.1/6
EducationYears of SchoolingNo household member of school entrance age +6 years has completed 6 or more years of schooling.1/6
EducationSchool AttendanceAny school-age child (up to the age of completing 8 years of schooling) not attending school up to grade 8.1/6
Standard of LivingCooking FuelHousehold uses dung, wood, charcoal, or coal for cooking.1/18
Standard of LivingSanitationHousehold uses a sanitation facility not shared with other households and not improved (per WHO/UNICEF guidelines).1/18
Standard of LivingDrinking WaterHousehold lacks access to improved drinking water or safe water is >30 minutes round-trip walk.1/18
Standard of LivingElectricityHousehold has no electricity.1/18
Standard of LivingHousingHousehold has at least one of: dirt, sand, dung, palm/bamboo, or similar floor; same for roof or walls.1/18
Standard of LivingAssetsHousehold owns fewer than 1 of: radio, TV, phone, bike, or motorbike; and no car or truck.1/18
A household is identified as multidimensionally poor if deprived in indicators weighted to at least 33% of the index (the poverty cutoff, or k ≥ 1/3), with a secondary intensity threshold applied such that the average deprivation among the poor also reaches at least 33% to emphasize acute rather than marginal poverty. This dual-cutoff structure aligns with the index's foundation in Amartya Sen's capability approach, which posits poverty as deprivations in essential functionings like nourishment and learning, rather than solely income shortfalls. However, the equal weighting across dimensions and sub-equal weighting within them remains a normative choice without robust behavioral or empirical derivation, rendering the framework susceptible to arbitrariness critiques; alternative data-driven weightings can alter rankings, though global MPI results show robustness to moderate perturbations.

Alkire-Foster Counting Method

The Alkire-Foster counting method aggregates multidimensional deprivations by first identifying the through a deprivation and then computing a product of poverty incidence and intensity, distinguishing it from additive or geometric aggregation techniques that may dilute information on the extent of overlapping deprivations. For each or individual, a deprivation score c_i is calculated as the weighted sum of deprivations across indicators, where a deprivation d_{ij} = 1 if the unit falls below the indicator-specific and 0 otherwise, and weights w_j reflect relative importance (often equally distributed within dimensions, summing to 1 overall). A unit is classified as multidimensionally if c_i \geq k, where k is a normative poverty representing the minimum weighted deprivations required (e.g., one-third of the total weighted indicators). Poverty incidence, or headcount H, measures the proportion of the identified as : H = \frac{q}{n}, with q as the number of poor units and n the total . A captures the average deprivation share among the poor: A = \frac{\sum_{i: c_i \geq k} c_i}{q}, reflecting the depth of beyond mere identification. The adjusted headcount , M_0 = H \times A, serves as the primary measure, satisfying axioms such as decomposability (allowing breakdown by subgroups or indicators), monotonicity ( rises with added deprivations), and robustness to dimensional among the non-poor. This counting approach enables precise targeting by decomposing M_0 into contributions from specific indicators or subgroups, revealing where interventions might reduce either incidence or most effectively. For instance, consider a of three households across three equally weighted indicators with k = 1/3. Household 1 is deprived in all three (c_1 = 1), Household 2 in two (c_2 = 2/3), and Household 3 in none (c_3 = 0). Households 1 and 2 are poor, yielding H = 2/3. The A = (1 + 2/3)/2 = 5/6 \approx 0.833, so M_0 = (2/3) \times (5/6) \approx 0.556. If Household 2 instead has deprivations totaling $2/3 (e.g., full deprivation in one and partial adjustment, but here), its intensity aligns directly with its score among the poor. Unlike Rawlsian metrics emphasizing the worst-off alone, the AF method balances breadth (via H) and depth (via A), avoiding undercounting moderate but widespread deprivations.

Computation and Dual Cutoffs

The computation of the Multidimensional Poverty Index (MPI) relies on the Alkire-Foster counting methodology, which applies dual cutoffs to identify deprivation and poverty at the level before aggregating to population-level measures. For each i, a deprivation score c_i is first calculated as the of weighted deprivations across the selected indicators: c_i = \sum_j w_j d_{ij}, where d_{ij} = 1 if the household is deprived in indicator j (based on its deprivation z_j) and $0 otherwise, and w_j are the normative weights (typically equal within dimensions, summing to 1 overall). The second cutoff, the poverty line k (commonly set at k = 0.333), determines multidimensional poverty: a household is poor if c_i \geq k, meaning it experiences deprivations whose weighted sum reaches at least one-third of the total possible. This dual-cutoff approach allows flexibility in parameter selection, with k reflecting a normative judgment on the breadth of deprivation required for poverty identification. The national MPI is then the population-weighted average of censored deprivation scores: M_0 = \frac{1}{N} \sum_{i=1}^N g_i(k), where g_i(k) = c_i if c_i \geq k and $0 otherwise, equivalent to the product of the incidence of H (share of with c_i \geq k) and the average intensity A (mean c_i among the poor). Computations incorporate survey design effects, such as clustering and stratification in Demographic and Health Surveys (DHS), using weighted estimators to ensure representativeness; for instance, Stata's svy commands adjust for primary sampling units and probability weights. Sensitivity to the poverty cutoff k is notable: increasing k (e.g., from 0.333 to 0.5) reduces H by excluding marginally poor households but raises A among the remaining poor, with the net effect on M_0 depending on the deprivation distribution; robustness tests by the Oxford Poverty and Human Development Initiative (OPHI) across k values like 0.2 and 0.4 often preserve country rankings, though headcounts vary substantially. Early implementations, including revisions, validated parameter choices through such analyses to enhance replicability across datasets.

Empirical Applications

Global and National MPI Estimates

The 2024 Global Multidimensional Poverty Index (MPI), jointly published by the Poverty and Human Development Initiative (OPHI) and the (UNDP), reports that 1.1 billion people across 112 countries—covering 6.3 billion individuals or 78% of the global population—experience acute , corresponding to an incidence rate (H) of approximately 18%. These estimates draw from harmonized household survey data spanning 2001–2023, with the MPI value (M0) reflecting both the headcount and average intensity of deprivations (A). Sub-Saharan Africa and South Asia dominate the global distribution, accounting for 83% of all multidimensionally poor people, with 522 million in the former and 402 million in the latter. In , incidence rates average close to 50%, varying widely by country but consistently higher than the global figure due to persistent deprivations in , , and schooling. South Asia exhibits lower average incidence, around 20–25% in recent years, though absolute numbers remain substantial amid population size and uneven progress across nations. National estimates reveal stark disparities and trends in . In , using comparable OPHI methodology, the multidimensional poverty headcount halved from 54.7% in 2005/06 to 27.5% in 2015/16, reducing the poor by 271 million, primarily through gains in access and cooking fuel. Government-adapted national MPI calculations show further decline to 11.28% by 2022–23, escaping 248 million more people since 2013–14. East Asian countries like demonstrate low national MPI values, with estimates around 5.5% incidence and 40.9% intensity based on domestic surveys, driven by advancements in living standards indicators such as and assets, though these differ from global harmonized data due to survey methodologies. Such variations highlight how national contexts influence MPI outcomes, with urban-rural divides and policy-targeted deprivations shaping distributions independently of monetary metrics.

Decompositions and Subgroup Analysis

The Alkire-Foster adjusted headcount ratio (M₀), the core measure of the Multidimensional Index, exhibits subgroup decomposability, whereby overall equals the population-share weighted sum of subgroup-specific M₀ values: M₀ = ∑_ℓ ν_ℓ M₀_ℓ, with ν_ℓ denoting the population share of ℓ. Each 's contribution is then D_ℓ⁰ = (ν_ℓ M₀_ℓ) / M₀, expressed as a share of total . This supports disaggregation by attributes such as rural-urban residence, head , , region, or age, quantifying disproportionate burdens—for instance, a comprising 25% of the population may contribute over 50% to national M₀ if its levels are markedly higher. Dimensional decomposition further breaks M₀ into weighted censored headcount ratios per : M₀ = ∑_k w_k h_k, where w_k is dimension k's weight and h_k is the proportion of the multidimensionally poor deprived in that . The intensity-adjusted contribution of dimension k is φ_k⁰ = (w_k h_k / M₀) × 100%, accounting for both deprivation prevalence among the poor and dimension weighting. This identifies primary drivers of breadth and depth; for example, in methodological illustrations, a like with h_k = 0.55 and w_k = 1/3 may contribute 20% to M₀, exceeding its nominal weight due to concentrated deprivations. Empirical applications leverage these techniques to pinpoint bottlenecks, though aggregation limits granular causal insights. Subgroup decompositions routinely reveal rural MPI values 2-3 times urban levels in low-income countries, while ethnic breakdowns in the 2021 global MPI across 40 countries exposed gaps exceeding 70% between advantaged and marginalized groups, such as indigenous populations facing compounded deprivations in and living standards. Dimensional analyses often highlight living standards (e.g., , ) as major contributors in and , or health (e.g., , ) in , guiding resource allocation despite challenges in disentangling correlated indicators.

Crisis Impacts and Case Studies

The disrupted multiple dimensions of , with simulations using pre-2020 household survey data from 70 countries estimating a reversal of 3.6 to 9.9 years of progress in global multidimensional . These projections, informed by UN data on closures and insecurity, highlighted spikes in deprivations—affecting attendance and years of schooling—and nutrition shortfalls due to supply chain breakdowns and income losses, dimensions often underemphasized in monetary assessments. Pre- and post-pandemic comparisons in affected regions showed MPI rising as households faced compounded deprivations, such as simultaneous access barriers from overwhelmed systems and living standard declines from asset sales. In , the National Multidimensional Poverty Index baseline from the 2015–2016 (NFHS-4) served as a reference, but subsequent NFHS-5 data (2019–2021) captured pandemic-era vulnerabilities, revealing stalled progress in urban and indicators amid lockdowns that closed schools for over 1.5 billion students globally, including 250 million in . Decompositions indicated that and weighted deprivations contributed disproportionately to any localized MPI upticks, underscoring the metric's utility in tracking non-monetary reversals without conflating them with causal policy effects. The 2010 Haiti earthquake provided an earlier case of crisis-induced MPI shifts, where post-disaster surveys documented heightened housing and asset deprivations, with decompositions attributing over 15% of the poverty intensity increase to destroyed infrastructure in a context of pre-existing high multidimensional poverty rates exceeding 50%. Such analyses, drawing on Alkire-Foster decompositions by indicator, illustrated how sudden shocks amplify living standards failures, distinct from gradual income erosions. For the 2022 , baseline MPI estimates stood at 0.2% (affecting about 100,000 people), but of over 5 million internally and damage projected escalations in utilities and deprivations, with early 2023 household signaling a 1.7-fold rise in broader vulnerability compared to 2021. These cases demonstrate the MPI's role in quantifying crisis resilience via pre/post indicator changes, capturing persistent non-monetary impacts like education disruptions that monetary metrics frequently miss.

Comparisons with Monetary Measures

Conceptual Differences

Monetary poverty measures, exemplified by the World Bank's international line of $2.15 per day in 2017 terms, gauge an individual's command over commodities through income or consumption thresholds. These metrics assume that sufficient resources enable the purchase of goods necessary for well-being, but they do not directly assess achieved outcomes or non-market deprivations. In contrast, the Multidimensional Poverty Index (MPI) focuses on deprivations in functionings—direct indicators of capabilities such as , years of schooling, and access to —rather than resource endowments. This distinction, rooted in Amartya Sen's capabilities approach, prioritizes what individuals can do or be over the means available to them, recognizing that resource possession does not guarantee valued outcomes due to conversion factors like personal heterogeneity or environmental barriers. A core conceptual divergence lies in aggregation: monetary measures implicitly permit trade-offs, where higher income might offset lacks in specific areas by enabling substitution. The MPI's Alkire-Foster counting , however, identifies households as poor if they experience deprivations across a sufficient weighted count of indicators (typically one-third), eschewing full compensation between dimensions during identification. For instance, MPI counts stunting as a deprivation irrespective of food spending, capturing nutritional failures that monetary lines overlook. This non-compensatory structure aligns with non-welfarist evaluations of absolute but has drawn critique for disregarding potential intertemporal or cross-dimensional substitutions observable in resource metrics. Debates underscore these foundations: Ravallion (2011) contends that multidimensional approaches add limited value beyond monetary , as deprivations often correlate strongly with income, rendering added dimensions redundant or arbitrarily weighted. Alkire and Foster (2011) rebut that such measures illuminate intrinsic deprivations in imperfect markets—such as access amid transfers—unrevealed by resource proxies, justifying multidimensionality for comprehensive appraisal despite empirical overlaps. These positions highlight tensions between resource-centric efficiency and outcome-oriented realism in conceptualization.

Empirical Divergences and Correlations

Empirical studies using household survey data demonstrate notable divergences between the Multidimensional Poverty Index (MPI) and monetary poverty measures at the country level. In , based on 2016 data, the monetary poverty headcount at the $1.90 international line was 33%, while the MPI incidence (at k=1/3) reached 51%, indicating a higher of multidimensional deprivations. Similarly, in using comparable data, monetary poverty affected 66% of the population at $1.90, compared to 76% under the MPI, with mismatches arising from deprivations in non-income domains such as and . Correlations between national MPI and monetary headcounts across 90 developing countries yield moderate Kendall coefficients of 0.641 to 0.719 (p<0.001), reflecting partial but imperfect alignment. These associations weaken in the poorest subgroups and countries, where multidimensional measures capture additional vulnerabilities not reflected in income or consumption thresholds. In advanced economies like , adapted MPI applications show low overall multidimensional —often below the 5% income rate (60% of )—yet persistent deprivations in subgroups such as low-education or immigrant households, highlighting "hidden" in assets, , or despite sufficient monetary resources. OPHI analyses of micro-data from six countries (, , , , , ) reveal substantial non-overlap in identified poor households, with 20-40% of those multidimensionally poor lacking corresponding monetary poverty in some cases, and vice versa; for instance, 22.2% in were only monetary poor, implying around 32% only MPI poor based on headcount differences. Such discrepancies underscore the MPI's sensitivity to overlapping deprivations, contrasting with monetary metrics' focus on expenditure shortfalls.

Policy Implications of Divergences

Divergences between multidimensional and monetary measures can lead to differing policy emphases, with the MPI advocating for targeted, sector-specific interventions to address non-income deprivations such as , , and schooling, while monetary metrics align more closely with broad strategies that have empirically driven large-scale reductions. For instance, MPI analyses often support direct subsidies or programs like improved in rural areas, aiming to alleviate overlapping deprivations simultaneously, whereas monetary-focused policies prioritize market liberalization and to boost incomes, as evidenced by China's post-1978 reforms that lifted approximately 800 million people out of through rapid GDP expansion and agricultural decollectivization by the early 2000s. Proponents of the MPI argue that its divergence from monetary measures enables more equitable targeting by identifying "hidden" pockets of missed by income thresholds, potentially enhancing social cohesion and long-term through multisectoral coordination. However, critics contend that over-reliance on MPI could divert resources from -oriented reforms, which have demonstrated stronger causal links to escape; empirical studies show that a 10% increase in GDP reduces multidimensional by only 4-5%, compared to near-proportional effects on monetary , suggesting targeted MPI interventions may underperform broad expansion without complementary . This elasticity gap implies risks of fiscal inefficiency, as prioritizing non-income dimensions might complicate budget allocation and yield slower aggregate reductions, particularly in contexts where market reforms have historically outperformed isolated sector programs. Evidence on whether MPI-guided policies surpass growth-focused alternatives remains mixed, with no robust causal demonstrations of superiority for direct interventions over liberalization-driven gains, underscoring the need for approaches that leverage 's proven scalability while using MPI for supplementary diagnostics. In resource-constrained settings, such divergences highlight a tension: while MPI promotes by spotlighting deprivations uncorrelated with , its may fragment efforts absent strong foundations, potentially hindering escapes from poverty traps observed in high-GDP contexts with persistent multidimensional gaps.

Criticisms and Limitations

Methodological Flaws

The Multidimensional Poverty Index (MPI), based on the Alkire-Foster counting methodology, assigns equal weights to each of its ten indicators grouped into three dimensions (, and living standards), with each dimension weighted at one-third overall. This equal weighting scheme lacks an empirical foundation and relies on normative judgments, as no data-driven method justifies treating all indicators or dimensions as equally important across diverse contexts. Sensitivity analyses demonstrate that altering these weights—such as emphasizing over or adjusting dimension shares—can significantly change country rankings and poverty headcounts, undermining the index's robustness for cross-national comparisons. Similarly, the deprivation cutoff of k=0.33, requiring at least one-third of weighted indicators to be deprived for identification as poor, is arbitrarily set without validation against principles or empirical deprivation patterns, leading to potential misclassification when thresholds shift. The MPI's reliance on binary deprivation cutoffs for each indicator introduces a discrete structure that violates the continuity axiom in poverty measurement, where small changes in should yield continuous adjustments in the index rather than abrupt shifts. For instance, marginal improvements in an indicator like years of schooling—such as advancing from 4.9 to 5.1 years—fail to reduce a household's deprivation count if it remains below the fixed of six years for two adults, resulting in no change to the MPI even as actual rises. This discontinuity also contravenes strict monotonicity, as the index does not monotonically decrease with every unit increase in any dimension unless it crosses a , contrasting with continuous measures that capture gradual progress. Furthermore, the MPI exhibits inequality blindness among the identified poor, as it averages deprivation intensity without penalizing uneven distributions of deprivations within the poor population. Transfers of deprivation shares from a severely deprived to one with milder shortfalls—such as reallocating access to assets—leave the overall unchanged, provided both remain above the k cutoff, unlike inequality-sensitive indices like the Foster-Greer-Thorbecke family applied to multidimensional settings, which incorporate progressive penalties for concentration among the poorest. This property reduces incentives for policies targeting the most deprived within the poor group, as the measure responds only to changes in incidence or average intensity, not dispersion.

Empirical Validity and Reliability Issues

The empirical validity of the Multidimensional Index (MPI) faces challenges when benchmarked against monetary measures, with studies showing moderate correlations (Spearman ρ ≈ 0.6) but poor concordance (Kendall's τ ≈ 0.4) between MPI and scorecards derived from . This discrepancy manifests in low (46.4%), where MPI identifies fewer households as poor compared to monetary thresholds (25.4% vs. 33.8%), potentially under-detecting deprivations aligned with income-based benchmarks. High specificity (85.4%) indicates MPI avoids false positives but misses substantial overlap, raising questions about its comprehensiveness as a poverty identifier. Reliability concerns arise from the MPI's dependence on cross-sectional household surveys such as Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Surveys (MICS), which create temporal and geographic gaps, limiting coverage to countries with recent data and excluding others. initiatives in 2024 and 2025 have standardized deprivation cutoffs and estimation methods across these surveys to enhance cross-country comparability, yet persistent proxy inaccuracies remain, including asset indices that imperfectly approximate household wealth and indicators subject to misclassification due to unmeasured causal factors like institutional access. For instance, revisions to the assets indicator have consolidated variables to better utilize available data, but empirical tests reveal ongoing limitations in capturing nuanced wealth disparities. In settings, MPI living standards indicators contribute less to overall (around 36% vs. 50% in rural areas), potentially understating deprivations by applying rural-oriented thresholds that overlook context-specific urban challenges, such as in dense environments. Revisions to indicators, incorporating natural materials in floors, roofs, and walls as deprivations, aim to address isolated cases but highlight verification difficulties, as proxy-based assessments struggle to disentangle conditions from broader causal dynamics without longitudinal data. These issues underscore the need for robust analyses to confirm MPI's alignment with ground realities across diverse contexts.

Policy and Causal Inference Concerns

The Multidimensional Poverty Index (MPI) has been critiqued for its limited utility in guiding policies that foster causal reductions in poverty, as it primarily identifies correlations among deprivations rather than mechanisms for behavioral or systemic change. Unlike monetary poverty measures, which decline in response to economic growth driven by incentives such as labor market participation and property rights, the MPI does not explicitly incorporate dimensions like employment status or economic freedoms that encourage individual agency and productivity. This omission can weaken its role in policy design, as evidenced by Martin Ravallion's 2010 analysis, which argued that the MPI's emphasis on slowly changing indicators like schooling and child mortality fails to reflect the rapid poverty-alleviating effects of pro-growth reforms, such as those implemented in China and India during economic upswings. Empirical evidence underscores the causal superiority of market-oriented growth in addressing both monetary and non-monetary poverty dimensions captured by the MPI. In and post-1990s, GDP per capita increases—often exceeding 6-8% annually in countries like and following —led to substantial declines in multidimensional deprivations, including improved , , and access, primarily through income-mediated channels rather than targeted non-monetary interventions alone. For instance, Vietnam's doi moi reforms from the mid-1980s onward halved monetary poverty rates by the 1990s while simultaneously reducing non-monetary indicators like child stunting, demonstrating that broad-based growth generates spillover effects across MPI dimensions more effectively than isolated service provisions. Critics contend this highlights a risk of policy misallocation under MPI guidance, where resources may prioritize direct inputs (e.g., subsidized assets) over outcome-oriented strategies like trade openness, potentially diverting attention from evidence-based growth policies that have proven causally effective. Debates around the MPI's policy implications, intensified in World Bank discussions circa 2010, reveal tensions between its advocates and skeptics. Proponents, including developers at the Oxford Poverty and Human Development Initiative, argue it enables granular targeting of overlapping deprivations, facilitating localized interventions that monetary metrics overlook. However, detractors like Ravallion warn it may distract from core reforms, as its aggregate structure lacks robust links to verifiable causal pathways for poverty escape, such as those validated in randomized evaluations of cash transfers or tied to markets. This perspective aligns with causal realism, prioritizing interventions with demonstrated general equilibrium effects over descriptive indices, though empirical validation remains contested given the MPI's relative novelty in policy evaluation frameworks.

Recent Developments and Future Directions

Updates to Indicators and Coverage

The 2025 Global Multidimensional Poverty Index (MPI) update expanded coverage to 109 countries, incorporating fresh survey data from 13 nations—including Azerbaijan (2023), Bangladesh (2022), and Bolivia (2023)—to encompass approximately 6.3 billion people. This builds on prior iterations, such as the 2024 report's 112 countries, by prioritizing recent household surveys for comparability while maintaining the core methodology of 10 weighted indicators across health, education, and living standards dimensions. Harmonized time series data, detailed in a 2025 methodological publication, now span 88 countries over 156 periods from 2001 to 2023/2024, allowing for standardized assessments of poverty incidence and intensity trends despite variations in national data collection frequencies. Disaggregations have been refined to reveal disparities by age groups (e.g., children under 18, working-age adults, elderly), rural/ residence, 1,359 subnational regions across 101 countries, and sex of household head, highlighting how multidimensional disproportionately affects vulnerable subgroups such as female-headed households and younger populations. These breakdowns, integrated into the 2025 report, support targeted without altering the household-level identification of , where individuals are deemed multidimensionally poor if deprived in at least one-third of weighted indicators. While the global MPI's indicators have remained consistent post-2010—encompassing deprivations in , , years of ing, attendance, cooking , , , , , and assets—national adaptations in select countries have incorporated bank account ownership as a within the assets to better capture . Methodological refinements address potential overlaps in and water indicators through fixed weightings (one-sixth each in living standards), ensuring additive contributions to deprivation scores rather than redundant counting, in response to earlier critiques on indicator interdependence. The 2025 OPHI-UNDP report aligns these elements with Sustainable Development Goal 1.2 by quantifying acute deprivations affecting an estimated 1.1 billion people, emphasizing empirical tracking over monetary thresholds. Projections for the Multidimensional Poverty Index (MPI) indicate that, based on pre-pandemic trends from onward, 51 out of 75 low- and middle-income countries were on track to halve their MPI incidence by 2030 through linear of historical rates. These models assume continuation of observed annual declines in the incidence and intensity of deprivations across , and living standards dimensions, without accounting for structural breaks. However, the introduced setbacks estimated at 3 to 10 years in timelines for affected countries, as simulated through adjustments to trend data reflecting increased deprivations in , , and schooling. Long-term trends reveal asymmetric progress across MPI dimensions, with deprivations in living standards—such as access to clean water, , , and cooking fuel—declining more rapidly than those in (nutrition and ) or (school attendance and years of schooling) in the majority of tracked countries. This pattern correlates empirically with public investments in , which enable scalable improvements in basic services, whereas and outcomes require sustained development and are more vulnerable to biological and demographic constraints. Among 86 countries with trend data spanning 2001 to 2023, 40 exhibited consistent MPI reductions, predominantly driven by living standards gains, underscoring the role of targeted capital expenditures over broader systemic reforms. These projections carry inherent limitations, as linear models overlook potential causal disruptions from reversals, geopolitical , or exogenous shocks, potentially overstating if stalls or understating it if innovations emerge. Emerging risks, including climate hazards, exacerbate vulnerabilities: nearly 80% of the 1.1 billion multidimensionally poor individuals in 2025 reside in regions exposed to high , , floods, or , which could compound deprivations and reverse trends through indirect effects on , , and . Thus, while historical patterns suggest feasible halving in on-track nations under conditions, real-world causal demands integrating elements for robust .

Ongoing Debates and Alternatives

Critics of the Multidimensional Poverty Index (MPI) contend that its deprivation cutoffs and equal overlook moderate deprivations affecting near-poor households, proposing instead the Moderate Multidimensional Poverty Index (MMPI), introduced in , which lowers the intensity threshold to capture at deprivation scores between 20% and 33% for targeted policy interventions. Proponents counter that extending MPI to moderate levels risks diluting focus on acute poverty, where empirical data show faster reductions under standard MPI tracking. A related controversy involves the MPI's aggregation method, which ignores positive correlations between dimensions like and , potentially understating compounded deprivations; alternatives such as the Correlation Sensitive Poverty Index (CSPI), developed from 2011 onward, incorporate these dependencies via adjustments, increasing sensitivity to among the poor by penalizing transfers that exacerbate disparities within poverty groups. Efforts to resolve foundational divides include preference-based unification models, as outlined by Decerf in , which derive poverty identification from welfare functions representing individual preferences, bridging monetary thresholds with non-monetary indicators to avoid dimension selection while aligning with revealed choice data over expert-assigned weights. Advocates for MPI highlight its value in exposing hidden non-monetary gaps uncorrelated with , such as sanitation access amid rising GDP, yet skeptics prioritize output-focused metrics like GDP for , noting cross-country regressions where a 10% GDP growth correlates with 4-5% MPI declines, attributing sustained primarily to market-driven expansion rather than multidimensional interventions whose impacts remain empirically ambiguous beyond growth effects. Critics further recommend hybrid approaches—monetary lines supplemented by targeted non-income monitors—over comprehensive indices, arguing the latter's arbitrariness in dimension inclusion undermines policy prioritization compared to verifiable economic drivers.

References

  1. [1]
    Global MPI | OPHI - Oxford Poverty and Human Development Initiative
    The global Multidimensional Poverty Index (global MPI) is an annual international measure of acute multidimensional poverty covering over 100 developing ...
  2. [2]
    Technical introduction to the AF Method | OPHI
    The Alkire-Foster (AF) Method developed by Sabina Alkire and James Foster offers an intuitive approach to the measurement of intrinsically multidimensional ...
  3. [3]
    [PDF] Technical note. Multidimensional Poverty Index
    The global Multidimensional Poverty Index (MPI) identifies multiple deprivations at the household level in health, education and standard of living. It uses.
  4. [4]
    A response to the weaknesses of the Multidimensional Poverty ...
    The specific structure of the MPI implies problematic distortions. It leads to an inflation in poverty rates that increases the poorer a country and thus the ...
  5. [5]
    [PDF] The multidimensional poverty index: Achievements, conceptual and ...
    Abstract. The Multidimensional Poverty Index (MPI) has been an important contribu- tion to the debate on national and international poverty measurement.Missing: peer | Show results with:peer
  6. [6]
    The Capability Approach - Stanford Encyclopedia of Philosophy
    Apr 14, 2011 · The capability approach is a theoretical framework that entails two normative claims: first, the claim that the freedom to achieve well-being is of primary ...Introducing the capability... · Core concepts and... · Specifying the capability...
  7. [7]
    [PDF] The Capability Approach and Multidimensional Poverty
    The capability approach defines poverty as a deprivation of capabilities, as a lack of multiple freedoms people value and have reason to value.
  8. [8]
    [PDF] Multidimensional Poverty Index - UN.org.
    Methodology: Alkire and Foster. The MPI combines two aspects of poverty. 1) Incidence ~ the percentage of people who are poor, or the headcount ratio H. 2 ...
  9. [9]
    Counting and Multidimensional Poverty Measurement | OPHI
    This paper proposes a new methodology for multidimensional poverty measurement consisting of an identification method ρκ that extends traditional approaches.
  10. [10]
    [PDF] Counting and Multidimensional Poverty Measurement
    Abstract. This paper proposes a new methodology for multidimensional poverty measurement consisting of an identification method ρκ that extends traditional ...
  11. [11]
    Counting and multidimensional poverty measurement - ScienceDirect
    This paper proposes a new methodology for multidimensional poverty measurement consisting of an identification method ρ k that extends the traditional ...
  12. [12]
    History of OPHI - Oxford Poverty and Human Development Initiative
    OPHI was founded in 2007 by Sabina Alkire and John Hammock, using the Alkire-Foster method, and developed the Multidimensional Poverty Index (MPI). It also ...
  13. [13]
    UN and Oxford University unveil new index to measure poverty
    Jul 14, 2010 · The new measure, the Multidimensional Poverty Index, or MPI, was developed and applied by the Oxford Poverty and Human Development Initiative (OPHI) with UNDP ...
  14. [14]
    Global MPI 2010 | OPHI
    Jul 14, 2010 · OPHI has developed a new international measure of poverty – the Multidimensional Poverty Index or MPI – for the 20th Anniversary edition.Missing: adoption | Show results with:adoption
  15. [15]
    Human Development Report 2010
    Jan 1, 2010 · This Report surveys critical aspects of human development, from political freedoms and empowerment to sustainability and human security.
  16. [16]
    [PDF] Human Development Research Paper 2010/11 Acute ...
    Jul 11, 2010 · This paper presents a new Multidimensional Poverty Index (MPI) for 104 developing countries. It is the first time multidimensional poverty ...
  17. [17]
    [PDF] Multidimensional Poverty Index
    In 2010, the 20th anniversary year of the United Nations Development. Programme's flagship Human. Development Report (UNDP HDR), the HDR is introducing a new.
  18. [18]
    Mexico Measure | OPHI
    Mexico's multidimensional poverty measure was made official in 2010 and has been updated regularly ever since. The National Council for the Evaluation of Social ...
  19. [19]
    [PDF] Methodology for Multidimensional Poverty Measurement in Mexico
    Sep 3, 2010 · Mexico is the first country whose normative framework created the legal basis for adopting a poverty measurement that recognizes its ...<|separator|>
  20. [20]
    Regional estimates of multidimensional poverty in India
    Results indicate that 43% of India's population are multidimensional poor with large regional variations. The average intensity of poverty was 45.5% with a MPI ...Missing: adaptation | Show results with:adaptation
  21. [21]
    Global MPI 2025 - Oxford Poverty and Human Development Initiative
    The 2025 global Multidimensional Poverty Index (MPI) uses the most recent comparable data available for 109 countries – 21 low-income countries, ...
  22. [22]
    2025 Global Multidimensional Poverty Index (MPI)
    Oct 17, 2025 · This 2025 Global Multidimensional Poverty Index (MPI) report, for the first time, overlays data on climate hazards and multidimensional ...1.1b · 740m · 887mMissing: OPHI partnership
  23. [23]
    MPI 2025 FAQs | Human Development Reports
    Since 2018 the MPI is therefore a contribution to the implementation and monitoring of Sustainable Development Goal 1 which aims to end poverty in all its ...
  24. [24]
    Global MPI 2024 | OPHI
    Nov 4, 2024 · The 2024 global Multidimensional Poverty Index (MPI) uses the most recent comparable data available for 112 countries—21 low-income countries, ...
  25. [25]
    The global Multidimensional Poverty Index: Harmonised level ...
    Jan 26, 2025 · The global MPI is an international poverty measure based on ten deprivation indicators in three dimensions: health, education, and living standards.
  26. [26]
    National Measures | MPPN - Multidimensional Poverty Peer Network
    Examples of official permanent national MPIs include: 2009: Mexico; 2010: Bhutan; 2011: Colombia; 2014: Ho Chi Minh City (Viet Nam) and South Africa (initial ...
  27. [27]
    [PDF] OPHI Country Briefing 2024: Colombia
    1 The global MPI is an index of acute multidimensional poverty that covers over 100 countries. It is computed using data from the most recent Demographic and ...Missing: adaptations | Show results with:adaptations
  28. [28]
    [PDF] Bhutan: Multidimensional Poverty Index 2022
    Bhutan's Moderate MPI is based on the data from the Bhutan Living. Standards Survey (BLSS) 2022, conducted by the National Statistics Bureau (NSB) funded by the ...
  29. [29]
    MPI and the SDGs | OPHI
    The MPI is a key tool to track progress towards SDG Target 1.2 and poverty reduction according to national definitions by 2030.Missing: monitoring MICS
  30. [30]
    [PDF] Improving the Availability & Use of Multidimensional Poverty Data
    Irregular or incomplete data hinders monitoring key goals like SDG 1. oAvailable data is often limited – frequency, modules, population coverage. oData ...
  31. [31]
    [PDF] Multidimensional Poverty Measurement and Analysis: Chapter 1
    We explore whether various unidimensional measures accurately reflect the level and trend of multidimensional poverty and related questions. We first probe ...
  32. [32]
    A debate on multidimensional poverty indices - World Bank Blogs
    Aug 3, 2010 · The more powerful critique of the MPI is that there is no solid basis for the choice of index weights. When I teach poverty measurement, I give ...
  33. [33]
    Is the Multidimensional Poverty Index Robust to Different Weights?
    How robust is this index to the different choices made in its design? This brief presents the results of tests that show that the MPI rankings are robust to a ...
  34. [34]
    [PDF] Chapter 5 – The Alkire-Foster Counting Methodology
    Computing the average share of weighted indicators in which poor people are deprived. This entails adding up the deprivation scores of the poor and dividing ...
  35. [35]
    Chapter 5 - Multidimensional Poverty Measurement & Analysis
    This chapter provides a systematic overview of the multidimensional measurement methodology of Alkire and Foster (2007, 2011a), with an emphasis on the first ...
  36. [36]
    [PDF] Multidimensional Poverty Index (MPI) - Human Development Reports
    The MPI follows the Alkire-Foster (AF) methodology developed by Sabina Alkire and James Foster. (2011). This measure assesses the simultaneous or “joint” ...
  37. [37]
    How is it Calculated? - MPPN
    The Alkire Foster methodology can be intuitively introduced in 12 steps. The first 6 steps are common to many multidimensional poverty measures; the remainder ...
  38. [38]
    [PDF] Testing the robustness of the revised multidimensional poverty index ...
    OPHI (2018) suggested testing the sensitivity of country rankings to the poverty cut- off of 33.33 per cent by using alternative cut-offs: 1 per cent, 20 per ...
  39. [39]
    (PDF) Sensitivity Analyses in Poverty Measurement: The Case of the ...
    Oct 20, 2020 · This paper provides an extensive sensitivity analyses of the global multidimensional poverty index (MPI), which is a counting-based measure of ...
  40. [40]
    2024 Global Multidimensional Poverty Index (MPI)
    Oct 17, 2024 · The results show that 1.1 billion of 6.3 billion people live in acute multidimensional poverty, over half of them children.MPI 2024 FAQs · MPI Country Notes · MPI Statistical Programmes
  41. [41]
    Global Launch of 2024 Multidimensional Poverty Index “Poverty ...
    Nov 13, 2024 · 83.2 percent live in Sub-Saharan Africa and South Asia, with respectively 522 million people and 402 million. Nearly two-thirds of poor ...Missing: incidence | Show results with:incidence
  42. [42]
    Global Multidimensional Poverty Index 2023
    Jan 29, 2024 · In Sub-Saharan Africa, poverty affects an average of 49.5 percent of the population, but incidence and MPI values vary widely across countries, ...<|control11|><|separator|>
  43. [43]
    Multidimensional Poverty Reduction in India 2005/6–2015/16
    Jul 15, 2020 · The paper found a strong reduction, halving the MPI, in multidimensional poverty in India from 2005/6 to 2015/16, with 271 million fewer people ...Missing: 2005-06 | Show results with:2005-06
  44. [44]
    271 million fewer poor people in India | United Nations Development ...
    Sep 20, 2018 · The incidence of multidimensional poverty has almost halved between 2005/6 and 2015/16, climbing down to 27.5 percent from 54.7 percent. Among ...
  45. [45]
    24.82 crore Indians escape Multidimensional Poverty in last 9 years.
    Jan 15, 2024 · 24.82 crore Indians escape Multidimensional Poverty in last 9 years. Steep decline in Poverty Headcount Ratio from 29.17% in 2013-14 to 11.28% in 2022-23.
  46. [46]
    [PDF] Exploring Multidimensional Poverty in China
    Overall, 5.5% of the Chinese population are MPI poor, and the average intensity of poverty is 40.9%. China's Global MPI value is 0.023 – making it similar to ...
  47. [47]
    Estimates of multidimensional poverty and its determinants among ...
    Oct 15, 2024 · This study focused on assessing the poverty status of rural older persons and providing a basis for decision-making at all levels of government in China.
  48. [48]
    [PDF] Unmasking disparities - by ethnicity, caste and gender
    For the first time the global MPI is disaggregated by ethnicity or race (for 40 countries with available information), by caste (for India) and by gender of ...Missing: decomposition | Show results with:decomposition
  49. [49]
    [PDF] Global Multidimensional Poverty Index 2023
    The global MPI is disaggregated to illuminate pockets of poverty and who is left behind. Finally, breakdown by component indicator shows what deprivations ...
  50. [50]
    Global multidimensional poverty and COVID-19: A decade of ...
    The COVID-19 pandemic and associated economic contraction are negatively impacting multiple dimensions of poverty and jeopardising this progress.
  51. [51]
    Global multidimensional poverty and COVID-19 - ScienceDirect.com
    We evaluate the potential impact of COVID-19 and policy responses on global poverty. •. We use 70 microdata sets underlying the global multidimensional poverty ...
  52. [52]
    [PDF] National Multidimensional Poverty Index - NITI Aayog
    Nov 24, 2021 · India's national MPI measure uses the globally accepted and robust methodology developed by the. Oxford Poverty and Human Development ...
  53. [53]
    2021 Global Multidimensional Poverty Index (MPI)
    Oct 7, 2021 · It explores how the pandemic has affected three key development indicators (social protection, livelihoods and school attendance), in ...
  54. [54]
    (PDF) Multidimensional child poverty in Haiti - ResearchGate
    The earthquake of January 2010 struck a country that was already one of the poorest in the world. This chapter shows that in the years before the earthquake ...<|separator|>
  55. [55]
    [PDF] Haiti in Distress: The Impact of the 2010 Earthquake on Citizen Lives ...
    Haiti shows a similar MPI poverty rate to that of Republic of. Congo and ... Multidimensional Poverty Index: 2010 Data. Oxford Poverty and Human ...
  56. [56]
    [PDF] Ukraine - Human Development Reports
    Based on these estimates, 0.2 percent of the population in Ukraine (100 thousand people in 2022) is multidimensionally poor while an additional 0.4 percent is ...
  57. [57]
  58. [58]
    Understandings and misunderstandings of multidimensional poverty ...
    Jun 18, 2011 · In this paper we elucidate the strengths, limitations, and misunderstandings of multidimensional poverty measurement in order to clarify the debate and ...
  59. [59]
    [PDF] Multidimensional Poverty Analysis: Looking for a Middle Ground
    Some of the debate between proponents and skeptics of multidimensional poverty measurement was featured in the Forum Section of the June 2011 issue of the ...
  60. [60]
    [PDF] Monetary and Multidimensional Poverty
    This paper examines the relationship between monetary and multidimensional poverty, finding a correlation that breaks down in poorer countries, and mismatches ...
  61. [61]
    Towards a Multidimensional Poverty Index for Germany | OPHI
    Sep 30, 2015 · His analysis of the MPI revealed substantive differences in the prevalence of multidimensional poverty among different groups of the ...Missing: divergence | Show results with:divergence
  62. [62]
    [PDF] Towards a Multidimensional Poverty Index for Germany
    Conceptually, I propose a more comprehensive operationalization of a MPI for an advanced economy like Germany, including a justification for selected dimensions ...
  63. [63]
    Policy uses of the MPI | OPHI
    Multidimensional poverty measures complement monetary poverty statistics by offering a wider perspective on poverty. They identify people who may be overlooked ...
  64. [64]
    Does economic growth reduce multidimensional poverty? Evidence ...
    The empirical analysis indicates that a 10% increase in GDP decreases multidimensional poverty by approximately 4–5%.
  65. [65]
    [PDF] Four Decades of Poverty Reduction in China - The World Bank
    Over the past 40 years, the number of people in China with incomes below US$1.90 per day has fallen by close to 800 million, accounting for close to three- ...
  66. [66]
    [PDF] China's Economic Growth and Poverty Reduction (1978-2002)
    [Abstract] This article has summarized and evaluated the trend of China's poverty decrease in the period of 1978-2002, considering that China has obtained ...
  67. [67]
    Multidimensional Poverty Measures as Relevant Policy Tools | OPHI
    This paper first situates muldimensional poverty measures in the Sustainable Development Goals, which seek to End Poverty in all its forms and dimensions.
  68. [68]
    Does economic growth reduce multidimensional poverty? Evidence ...
    Our results indicate that economic growth is an important instrument to alleviate multidimensional poverty, but its effect is substantially lower than that on ...
  69. [69]
    [PDF] 1 Growth and Poverty Revisited from a Multidimensional Perspective ...
    Results from a First Difference Estimator Model suggest that while economic growth reduces multidimensional poverty, this impact is well below a one-to-one ...
  70. [70]
    understanding poverty through the Multidimensional Poverty Index
    Dec 16, 2024 · The MPI measures how people experience poverty in key areas of well-being across more than 100 countries.
  71. [71]
    [PDF] The Multidimensional Poverty Index: Achievements, Conceptual and ...
    After determining the indicator cut-offs, the Alkire-Foster method attaches weights to each deprivation. The MPI weighs each dimension equally (1/3) and within ...
  72. [72]
    [PDF] Multidimensional poverty indices: A critical assessment
    Jul 10, 2014 · 3.1 Continuity. It is clear from Figures 6 and 7 that headcount-type multidimensional indices cannot obey the continuity axiom. This is ...
  73. [73]
    Assessing Trends in Multidimensional Poverty During the MDGs
    Mar 9, 2022 · 1 Second, the MPI is insensitive to inequality among the poor, which is an important property that every poverty index should have (Jenkins ...
  74. [74]
    [PDF] Monitoring progress in multi-dimensional poverty reduction - EconStor
    Secondly, the M measure pays no attention to the distribution of deprivations; it is thus totally insensitive to inequality among the multi-dimensionally poor ...<|control11|><|separator|>
  75. [75]
    (PDF) Measuring and Decomposing Inequality among the ...
    A poverty measure that is insensitive to inequality does not provide incentives to the policymaker to prioritize the conditions of the poorest. ... Determinants ...
  76. [76]
    Reliability and validity of multidimensional poverty index with ... - NIH
    Sep 24, 2025 · Reliability and validity of multidimensional poverty index with poverty scorecard in measuring household poverty in rural India. Nafis Faizi ...
  77. [77]
    Consolidating and improving the assets indicator in the global ...
    The article validates a revision to the assets indicator of the global Multidimensional Poverty Index (global MPI) that makes the best possible use of existing ...
  78. [78]
    Full article: Revisions of the global multidimensional poverty index
    Dec 15, 2020 · Following this argument, the revised global MPI considers natural and rudimentary roofs and walls, and natural flooring to be deprived. We ...Missing: causality | Show results with:causality
  79. [79]
    World Bank research director critiques the new UN poverty index
    Jul 28, 2010 · The MPI would probably not capture well the impacts on poor people of economic downturns (such as the Global Financial Crisis) or rapid upswings ...
  80. [80]
    Do Monetary and Nonmonetary Indicators Tell the Same Story About ...
    This paper investigates whether monetary and nonmonetary indicators tell the same story about chronic poverty using a unique panel data from Vietnam in the ...
  81. [81]
    [PDF] – A Review of MPI Measurement and Uses in Asia and the Pacific
    Jul 18, 2019 · It recommended using a global MPI to complement measures of income poverty that are based on US$1.90 a day purchasing power parity (PPP). Second ...
  82. [82]
    The Multidimensional Poverty Index Debate: rounds 2, 3, 4 …
    Aug 10, 2010 · There has been a first-class debate on various blogs on the Multidimensional Poverty Index (MPI), including here, in this blog. Here's a brief round-up.
  83. [83]
    [PDF] MULTIDIMENSIONAL POVERTY INDEX 2025
    Oct 16, 2025 · This 2025 Global Multidimensional Poverty Index (MPI) report, for the first time, overlays data on climate hazards and multidimensional poverty ...Missing: partnership | Show results with:partnership
  84. [84]
    [PDF] MULTIDIMENSIONAL POVERTY INDEX 2024
    The global MPI is a key international resource that measures acute multidimensional poverty across more than 100 developing countries (box 1). First launched in ...
  85. [85]
    The global Multidimensional Poverty Index (MPI) 2025
    ... gender of household head) of the global Multidimensional Poverty Index (MPI) 2025. The 2025 MPI disaggregation results are based on the most recent data ...
  86. [86]
    [PDF] The Global Multidimensional Poverty Index (MPI) 2022
    The global MPI tables identifies individuals as vulnerable to poverty if they are close to the one-third threshold, that is, if they are deprived in 20 to 33. ...Missing: ethnic | Show results with:ethnic
  87. [87]
    On Track or Not? Projecting the Global Multidimensional Poverty Index
    If trends continue, 47 countries will halve poverty by 2030, but COVID-19 may cause a 3-10 year setback. 51 countries were on track before.
  88. [88]
    On track or not? Projecting the global Multidimensional Poverty Index
    We propose a framework for modelling projections of multidimensional poverty. We find that 51 of 75 countries are on track to halve poverty incidence by 2030.
  89. [89]
    New Global Multidimensional Poverty Index Report Reveals Nearly ...
    Oct 17, 2025 · Nearly 8 in 10 people living in multidimensional poverty – 887 million out of 1.1 billion globally – are directly exposed to climate hazards ...
  90. [90]
    Moderate Multidimensional Poverty Index: Paving the Way Out of ...
    It was developed to measure 'acute' multidimensional poverty globally and reflect aspirations in the Millennium Development Goals (MDGs).
  91. [91]
    [PDF] The Correlation Sensitive Poverty Index (CSPI)
    The CSPI captures inequality among the poor; it in- creases whenever items are transferred from a poor to a less poor household.
  92. [92]
    An Individual-based Index of Multidimensional Poverty for Low
    Aug 27, 2021 · This paper proposes a new index of multidimensional poverty, called the Global Correlation Sensitive Poverty Index (G-CSPI), which has three interesting ...
  93. [93]
    A preference-based theory unifying monetary and non-monetary ...
    Our theory makes no conceptual distinction between a “multidimensional structure” and a “unidimensional structure” to poverty measurement (Chakravarty and Lugo ...
  94. [94]
    [PDF] Growth and Poverty Revisited from a Multidimensional Perspective
    In Equation (10) the rate of poverty reduction is regressed on the rate of growth in mean GDP per capita and thus 𝛽 can be directly interpreted as the growth ...