Wealth indices created separately for different areas within a dataset using the same set of indicators
| Study (first author) | Dataset | Wealth index–consumption expenditure association |
| Jamal40 | Pakistan Integrated Household Survey 2001/2 | R2 values from regressions of indices on consumption expenditure: |
| Urban: 0.69 | ||
| Rural: 0.52 | ||
| Lindelow47 | Mozambique National Household Survey on Living Conditions | Re-ranking: urban households ranked higher by wealth index than consumption expenditures; vice versa for rural households. Poorer, more remote areas also lose rank even after controlling for urban/rural residence |
| Montgomery38 | Ghana 1988 (LSMS) | R2 values from regressions of indices on consumption expenditure: |
| Whole population: 0.104 | ||
| Urban: 0.082 | ||
| Rural: 0.014 | ||
| Montgomery38 | Jamaica 1989 (LSMS) | R2 values from regressions of indices on consumption expenditure: |
| Whole population: 0.143 | ||
| Urban: 0.094 | ||
| Rural: 0.106 | ||
| Montgomery38 | Pakistan 1991 (LSMS) | R2 values from regressions of indices on consumption expenditure: |
| Whole population: 0.030 | ||
| Urban: 0.036 | ||
| Rural: 0.025 | ||
| Montgomery38 | Peru 1994 (LSMS) | R2 values from regressions of indices on consumption expenditure: |
| Whole population: 0.154 | ||
| Urban: 0.108 | ||
| Rural: 0.132 | ||
| Montgomery38 | Tanzania 1993/4 (LSMS) | R2 values from regressions of indices on consumption expenditure: |
| Whole population: 0.155 | ||
| Urban: 0.114 | ||
| Rural: 0.017 | ||
| Skoufias42 | Mexico 1996 ENIGH | Sensitivity: |
| Urban: 53.4% | ||
| Rural: 67.6% | ||
| Ward21 | Tanzanian Household Budget Survey 2000/01 | % households in correct tercile: |
| Dar es Salaam: 60.4% | ||
| Urban: 63.5% | ||
| Rural: 58.6% |
LSMS, Living Standards Measurement Survey.