Absolute poverty measures a fixed threshold of resources needed for basic subsistence, while relative poverty measures falling below a share of a society's median income or standard of living. The two definitions can move in opposite directions: economic growth can lower absolute poverty while leaving relative poverty unchanged if inequality persists.
A poverty line is not a natural fact but a deliberate statistical construction, often anchored to the cost of a minimum food basket multiplied by a factor for non-food needs. Because the choice of basket, multiplier, and adjustment reflects judgment, the measured poverty rate shifts with these methodological decisions rather than with circumstances alone.
Poverty can be understood as a deprivation of capabilities, the real freedoms a person has to achieve valued states of being and doing, rather than only a shortfall of income. Under this view two people with identical incomes can face different poverty because conversion of income into functioning depends on health, disability, local prices, and social context.
Below certain thresholds, low resources reproduce low resources: insufficient nutrition reduces earning capacity, lack of collateral blocks credit, and poor households cannot afford the lumpy investments that would raise future income. These mechanisms can hold households or regions in a stable low state absent an external push large enough to clear the threshold.
Viewing poverty through multiple simultaneous dimensions, education, health, housing, sanitation, and income, reveals households that a single income measure would miss or misclassify. A multidimensional index counts the overlapping deprivations a person actually experiences, changing both who is identified as poor and which interventions appear most urgent.
Transfer programs must balance coverage against leakage: tightly targeting the poor reduces cost but risks excluding eligible people and creating disincentives at the eligibility cutoff, while universal programs reach everyone but spend heavily on the non-poor. The design choice is therefore a tradeoff among accuracy, administrative cost, and work incentives rather than a single optimum.