Inequality and Poverty Lecture Notes PDF

Inequality and Poverty Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Context - authors investigate the relative effectiveness of food program targeted at poor households in Indonesia...

Inequality and Poverty Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Context - authors investigate the relative effectiveness of food program targeted at poor households in Indonesia - in 2017 Indonesian government began reform of its largest anti- poverty program from in-kind to voucher program - in-kind food assistance program: 10 kg of free rice per month - electronic voucher program: debit card for buying similar value of rice and eggs - authors exploit randomization in program reform to analyze relative program effectiveness Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Theory - in-kind program : increase demand and supply of rice (ambiguous price effect) - voucher program : increase demand for rice, eggs (price increases) Program Administrative Challenges - in kind program: cumbersome/costly direct distribution of goods - voucher program: technology barriers (ex: cell phone reception for use of debit machine) Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Research Design - Indonesian government program Rastra (since 1998) provides subsidized food (rice: 10kg) to 15 million households per month with a budget of $1.5 billion USD - organized by central logistics agency, distributed by local government - in 2017 Rastra replaced by voucher program Non-Cash Food Aid (debit card for approximate value of Rastra rice: Rp 110,000, to be used for rice and eggs at designated shops) Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Research Design - Rastra replacement was conducted in phases starting in 2017 - Government + authors randomized treatment of 105 districts in 2018 due to budget limitations (42 treated in 2018, 63 scheduled for 2019, with additional seasonal randomization of treated districts) - Authors use these 105 districts as their sample in their analysis - These districts had a total population of 53 million with 3.4 million tragetet beneficiaries Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Data - SUSENAS (national survey) March 2018 and March 2019: special questions on whether household received Rastra or Voucher program + details of benefits - Unified Targeting Database (government database used to identify program eligibility): baseline household characteristics from survey data - 2018 village census: village-level baseline controls Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Estimation Strategy - y: outcome variable (value of subsidy received, any subsidy, etc.) - voucher: treatment group - X: household controls - h: household, v: village, s: strata, d: district - alpha: strata fixed effects Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Key Results Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Key Results lower PMT score indicates poorer household Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 targeted households PMT30 received Rp 9,162 on averagein in- kind program Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Key Results Subsidy value vs contemporaneous consumption (instead of PMT which is based on previous years’ data) Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414./514 Poverty reductions from voucher vs in-kind program Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414./514 Summary of key results - similar amount of aid distributed across programs, but voucher program households received subsidy worth 85 percent more, conditional of receiving a benefit - poorer voucher households were 16 percent less likely to receive benefits vs in-kind but richer voucher households were 49 percent less likely to receive benefits vs in-kind => less households received benefits - BUT poorer voucher households received 46 percent MORE benefits compared to in-kind, leading to 4.6 percentage points decrease in poverty for poorest 15 percent of households Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414./514 Mechanisms - the authors investigate the difference in the amounts of benefits received between the voucher and in-kind programs: - in-kind benefits are divisible (rice is packaged at the local level); vouchers are indivisible. Government agents decide how much rice to allocate to each individual - in-kind benefits distributed by local government officials; voucher benefits distributed by banks and private shops. Private agents less vulnerable to political pressure Banerjee et al. Electronic Food Vouchers: Evidence from an At-Scale Experiment in Indonesia commented by Elise Critoph for the course ECON 414/514 Conclusions - authors examine results of experiment on in-kind versus voucher assistance programs - poor and near poor areas in voucher program received 46 percent more assistance compared to in-kind program. - poverty rate was reduced by 20 for households in the bottom 15 percent - some small price increases in rural areas from voucher over in-kind outweighed by benefits of program - voucher program is cheaper to run -> administrative gains - voucher households consumed higher quality rice than in-kind household - voucher household increased their consumption of eggs relative to in-kind households - authors’ results suggest that a voucher program over in-kind distribution can reduce poverty Measuring Inequality Personal or size distributions Total incomes received by individual or households Source, location and occupational source of income is ignored. Total population is divided into groups/sizes Determine the proportion of income received by each group Kuznets ratio: a measure of the degree of inequality between high and low income groups in a country (ratio of income received by top 20% and bottom 40% of population) Table 5.1 Typical Size Distribution of Personal Income in a Developing Country by Income Shares—Quintiles and Deciles Measuring Inequality Lorenz Curves Horizontal axis: income recipients in cumulative percentages Vertical axis: share of total income received Measuring Inequality Lorenz Curves Perfect equality Measuring Inequality Lorenz Curves Perfect equality Measuring Inequality Lorenz Criterion When one curve lies above the other curve, the former is more equal than the latter. Measuring Inequality Gini Coefficients The area between the diagonal and the Lorenz curve divided by the total area of triangle in which the curve lies. An aggregate measure of inequality between 0 and 1. Measuring Inequality Gini Coefficient Properties 1. Anonymity principle: the measure does not depend on who has the higher income 2. Scale independence principle: the measure does not depend on the size of the economy or currency used 3. Population independence principle: the measure is not based on the number of income recipients 4. Transfer principle: if some income is transferred from a rich person to a poor person (not so much that they are now richer than the rich person), the new income distribution is more equal. Measuring Absolute Poverty Income Poverty The number of people who are unable to command sufficient resources to satisfy basic needs. The total number “headcount” ,H, whose incomes are below the absolute poverty line, Yp. Total poverty gap (TPG) measures the total income necessary to raise those below the poverty line, up to the poverty line. Measuring Absolute Poverty Income Poverty Total poverty gap (TPG) Measuring Absolute Poverty Income Poverty Total poverty gap (TPG) Measuring Absolute Poverty Income Poverty Total poverty gap (TPG) TPG = Average poverty gap (APG) APG = Measuring Absolute Poverty Income Poverty Normalized poverty gap (NPG) NPG = Kuznets’s Inverted-U hypothesis In the early stages of economic growth, the distribution of income worsens and then improves at later stages. The Kuznets curve reflects the relationship between a country’s income per capita and its inequality of income Growth and Poverty Reducing poverty need not lead to slower growth 1. Widespread poverty creates conditions in which the poor have no access to credit 2. The rich in many poor countries do not save and invest substantial proportions of their incomes 3. Low incomes and low levels of living can lower productivity, leading to slower growth 4. Raising incomes of the poor will increase demand for locally produced necessities 5. A reduction of mass poverty acts as a powerful incentive to participate in the development process