Development Economics Lecture 5: Poverty Alleviation PDF

Development Economics Lecture 5: Poverty Alleviation Lukas Hensel Peking University October 17, 2024 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Nobel Prize 2024...

Development Economics Lecture 5: Poverty Alleviation Lukas Hensel Peking University October 17, 2024 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Nobel Prize 2024 Development Economics 1 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Discussion online Development Economics 2 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Housekeeping Last week’s lecture is available online. Please check it out and ask questions if anything is unclear. Reminder: The first problem set is uploaded on the course page. It is due on Thursday the 24th of October. No presentations today but starting from next week we will have plenty. Development Economics 3 Poverty Trends Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Absolute poverty over time Absolute poverty declined (almost) everywhere. In 1990 about 38% of the world population lived in absolute poverty... In 2019 it was less than 9% of the world population. Development Economics 4 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions This also holds for higher poverty thresholds ! Progress here is more heterogeneous. Development Economics 5 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions headcounts have declined similarly But 650 million people still live in absolute poverty. Development Economics 6 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Recent progress in headcounts even for higher thresholds China’s middle class really makes a di↵erence here! Development Economics 7 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions A key driver of poverty reduction was economic growth. ! We will discuss trade and the quality of institutions as drivers of economic growth later on. Today, we focus on direct poverty alleviation policies for the people remaining in poverty. ! We are tinkering with the details of poverty alleviation programs. First, we need to understand what living in extreme poverty means. ! Development Economics 8 Living in Poverty Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The development economics power couple: Esther Duflo and Abhijit Banerjee They were at the heart of the credibility revolution of (development economics). Both are incredibly productive and you will see many more of their papers throughout. Duflo was only the second woman and the youngest awardee to receive the prize. Both received the Nobel Prize in Economics in 2019 for their work in development economics. Development Economics 9 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The Economic Lives of the Poor 1. What was the most interesting/important thing you learned? 2. What was the most puzzling/surprising thing you learned? The economic choices of the poor are constrained by their market environment. a safe place to put their savings. lack of shared infrastructure. When the government builds a water line to your neighborhood, for example, you no longer Development Economics 10 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Prominent household survey datasets Living Standards and Measurement Surveys (LSMS) 1. World Bank’s flagship measurement project since 1985. Basis for many other surveys. 2. Some of the earliest comprehensive, nationally representative datasets. 3. 110 surveys in 37 countries. 4. Originally cross-sectional surveys, now sometimes longitudinal/panel surveys. Demographic and Health Surveys (DHS) Focused on “population, health, HIV, and nutrition” data >400 surveys in 91 countries Cross-sectional surveys Family Life Surveys 1. Indonesia, Malaysia, Mexico 2. Typically panel surveys (4 waves over 22 years in Indonesia) 3. What does “family survey” imply relative to “household survey”? Development Economics 11 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Key economic patterns to recall and/or explain Defining the poor 1. Use consumption thresholds 2. Assume poverty is permanent 3. Acknowledge within-country price adjustment is difficult Living arrangements 1. Large households, to spread fixed costs 2. Many young people, possibly due to high mortality Earning 1. Widespread self-employment, but very small scale and not growing 2. Low specialization, probably limits earnings but reduces risk 3. Widespread temporary but not permanent migration Development Economics 12 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Key economic patterns to recall and/or explain Spending 1. Spend 56-78% of income on food consumption, don’t seem to be maximizing calories, also want higher quality calories (acknowledge overconsumption is becoming an issue in some middle-income countries) 2. Large spending on alcohol, tobacco, and festivals; very little on ‘developed-world-style’ entertainment Markets 1. Little formal credit access, expensive to use formal credit, may reflect high screening costs or market power (lots of informal credit use) 2. Few formal savings accounts, many informal: SHGs, ROSCAs, now microcredit 3. Little formal insurance, low take-up of formal insurance when it’s o↵ered, widespread informal insurance but education & health e↵ects of shocks indicate insurance is not complete 4. Bad land markets with insecure titles Development Economics 13 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Key economic patterns to recall and/or explain Infrastructure 1. Wide variation in access to electricity, water, sanitation, etc. 2. Generally better in urban than rural areas 3. Reasonable access to education & health services but the quality is terrible, hence bad health outcomes and low levels of learning 4. Some evidence private sector is growing to meet the demand for quality These are based on data from 1998 to 2005. How do to think the lives of the poor developed over the last 15 years? Development Economics 14 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions What can policy makers do to alleviate poverty? Development Economics 15 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions What are the objectives of poverty alleviation? Maximize ”total welfare” of the relevant population Reduce poverty today and in the future Take welfare of future generations into account Provide insurance against negative shocks They are subject to constraints Budget constraint (loans, tax revenue, grants) Political constraints ) Constraint optimization (i.e. cost-e↵ectiveness is important) Development Economics 16 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Particularly valuable types of intervention Improve outcomes of young generations: Beneficial from a utilitarian perspective (benefits last over a life-cycle) Investments are intertemporal complements (hence return to early life investments greater than later investments) ! They are also more likely to pay for themselves One-o↵ interventions that permanently lift people out of poverty More fiscally sustainable (and hence scalable) Empowers recipients Feasibility unclear Development Economics 17 Poverty traps Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions When can one-o↵ interventions permanently reduce poverty? There are two possibilities: 1. Shift the current steady state of the individual 2. Shift the individual to a new, higher steady state When individuals (or countries) are stuck in a relatively low equilibrium we speak of ”poverty traps.” Development Economics 18 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Model set-up Let’s sketch a theory of poverty traps for individual i (in parallel to the Solow model). Ki,t+1 = i (Ki,t ) = si Ai f (Ki,t ) + (1 )Ki,t (1) Steady state: Ki⇤ = ⇤ i (Ki ) (2) Development Economics 19 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Di↵erent individuals can have di↵erent steady states Kt+1 45 Kj⇤ j (Kjt ) i (Kit ) Ki⇤ Kt 0 Ki⇤ Kj⇤ ) but this does not mean that individuals face a poverty trap. Development Economics 20 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions S-shaped capital accumulation can lead to poverty traps Kt+1 45 (Kt ) Kh⇤ ⇤ Km Kl⇤ Kt 0 Kl⇤ ⇤ Km Kh⇤ Development Economics 21 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions How can such S-shaped arise? Such a shape could, for example, arise due to increasing returns to scale in f (·) or if si is an increasing function of Ki,t. What are practical examples? Development Economics 22 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions How can poverty traps arise? More generally, poverty traps can arise for a range of reasons: Production-based poverty traps (returns to scale, discontinuities) Human capital poverty traps (nutrition, education, health) Regulatory poverty traps (migratory and occupation restrictions) Financial poverty traps (lack of credit, lumpy investments, self-control issues) Psychological poverty traps (aspirations, stress, mental health) Lack of aggregate demand (big push policies) - needs coordination across actors Development Economics 23 Modern poverty alleviation programs Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Two kinds of asset transfer programs Multifaceted poverty alleviation Cash transfers (Haushofer and Shapiro, program (Banerjee et al., 2015) 2016) Often agricultural assets (cows, goats, Can be both unconditional and chicken). conditional (on school attendance, Often accompanied by training and health check-ups). ongoing support services. Governments often o↵er repeated Many have a small cash component smaller payments while NGOs often o↵er large, one-o↵ payments. Often transferred by mobile money to avoid leakage Development Economics 24 Multifaceted poverty alleviation programs Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Banerjee et al. (2015) This study tests the e↵ectiveness of complex, multifaceted poverty alleviation programs across six countries. The program is modeled after a successful program developed by BRAC, one of the biggest and most successful NGOs in that space. BRAC started in Bangladesh in 1972 and now operates in 12 countries worldwide. They use rigorous impact evaluation throughout to test their programs. They now on many sectors and interventions beyond poverty graduation programs. Development Economics 25 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Treatment components What were the di↵erent components of the treatment? Development Economics 26 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Randomization Many two-stage randomizations to study spillover e↵ects. Public lotteries can help build trust and avoid backlash. Development Economics 27 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Aside: Two-stage randomization With two-stage randomization we get the following data structure: Control village Treatment Village Control household YCC YT C Treatment household - YT T ) YT C YCC is an estimate of spillover e↵ects of the intervention on non-treated eligible households. What about ineligible households? ) We can compare ineligible households in treatment and control villages as an additional measure of spillovers. Development Economics 28 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Study timeline Development Economics 29 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Analysis Yk = ↵ + 1 assignmenti + Zik + Tcountry + Ushortsurvey + Vstratif ication + "i (3) ) This includes baseline value of the outcomes (Zik ), country dummies (Tcountry ), survey version dummies (Ushortsurvey ), and stratification variables (Vstratif ication ). Every observation is weighted equally ! What is the implication of this? Development Economics 30 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Results after three years (endline 2) All outcomes were standardized to make them comparable across sites: Yi mean(Y ) yisd = sd(Y ) We see mostly positive e↵ects of the interventions! Development Economics 31 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions MHT corrected results When combined into indexes the results remain highly significant. Are these large e↵ect sizes? ! For standard normal variables, 0.1 SD is moving from the median to the 54th percentile (or from the 10th percentile to the 12th ). Correcting for multiple hypothesis testing across indexes with q-values does not change the results. Development Economics 32 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The program was, on average, highly cost-e↵ectiveness What can we infer from positive rates of return to this intervention? Development Economics 33 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The implication of positive rates of returns Does this suggest that there is a market failure? If there is such a big return why don’t households borrow to obtain the same services? There are several possibilities: Development Economics 34 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Do we learn anything about the existence of poverty traps Consumption and assets remain higher after the end of the treatment. How can we interpret this? ! We cannot distinguish between a new steady state and escaping a poverty trap. The intervention explicitly aimed to change the savings rate ! this suggests a new steady state. The interventions include the transfer of large, productive assets ! this suggests a poverty trap story. Development Economics 35 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Remember the two possibilities a) Two steady states b) Poverty trap Kt+1 Kt+1 45 45 0 Ki⇤ i Kit Kh⇤ (Kt ) i (Kit ) Ki⇤ Kl⇤ Kt Kt 0 Ki⇤ Ki⇤ 0 Kl⇤ Kh⇤ Development Economics 36 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions How can we test whether poverty traps exist? Balboni et al. (2021) test empirically whether poverty traps exist. They study a large productive asset transfer (mostly a single cow) in Bangladesh with 6000 poor households. They also collect data on richer households to describe the whole distribution of outcomes. Development Economics 37 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Bimodal distribution of capital (in 1000 Bangladeshi Taka) at baseline This is necessary but not sufficient for poverty traps to exist. Development Economics 38 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The transfer creates a bump between the two modes of the distribution Where do these individuals move to from there? Development Economics 39 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Estimating the transition equation They estimate the capital transition equation from 2007 to 2011 and find an S-shape. Development Economics 40 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Interpretation of the S-shape This indicates that there is a capital threshold above which individuals converge to the new steady state. The asset transfer is worth around 490 USD and the estimated threshold is 504 USD. ! The di↵erence of 14 dollars is roughly worth a plow or a cart to make the cow (more) productive. This threshold falls right into the ‘bump’ of the post-transfer asset distribution. Whether people are moved to just below or just above the threshold depends on their initial capital stock. Development Economics 41 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Identification challenge To compare individuals who were pushed above and below the threshold, they would need to assume independence of the initial capital stock from the potential impact of the potential outcomes. Why might this not hold? They use a di↵erence-in-di↵erences estimator with the control group (more on this soon). ! For now just think of them as treatment e↵ects in each year. Development Economics 42 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Long-term results (di↵erence within treatment group by I(ki,2007 > k̂)) a) Assets b) Consumption Households above the threshold have Note that consumption first decreases and consistently more assets than those below. then increases! Development Economics 43 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Take-away Both descriptive patterns and treatment e↵ects on capital accumulation point towards the existence of a poverty trap. They follow treatment households above and below the poverty threshold over 11 years and find that the two groups’s consumption patterns diverge over time (in line with a poverty trap model). What are the implications for policymakers? A slight increase in the transfer size could have disproportional positive e↵ects. But anticipating such thresholds is very tricky. Development Economics 44 Cash transfers Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Cash transfers as poverty alleviation The concept is simple: give poor people cash. ! mobile money makes this secure and cheap! What are the advantages relative to the multifaceted interventions? What are the disadvantages relative to the multifaceted interventions? Development Economics 45 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Misspent money? What do you think, how will people spend the money? Development Economics 46 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Conditional vs unconditional cash transfer programs Conditional cash transfers Unconditional cash transfers Tied to school attendance, health Money has no strings attached. check-ups, vaccinations, etc. Mainly promoted by Give Directly (NGO) Used by many governments to mitigate relying on philanthropic funds (related to extreme poverty (famous examples: Mexico the e↵ective altruism movement). (Opportunidades); Brazil (Bolsa Familia); Popular with many private donors because Indonesia (PKH)). of low overhead cost. Promoted by the World Bank as ‘best practice’. Fundamentally they have similar goals: reduce poverty by giving cash. Development Economics 47 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Give Directly (https://www.givedirectly.org) Founded in the US in 2008 by PhD students at MIT and Harvard. Has given away more than 700 million dollars to poor individuals in Bangladesh, DRC, Kenya, Liberia, Malawi, Mozambique, Morocco, Nigeria, Rwanda, Turkey, Uganda, USA, and Yemen. They are very much guided by RCT evidence. ! They have many ongoing experiments, including a universal basic income trial going on. Development Economics 48 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions How would you design an RCT to study cash transfers? Study parameters: What sample? Level of randomization? Data collection? Development Economics 49 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The sample Who would you include in the study? Development Economics 50 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Level of randomization At what level would you randomize? Development Economics 51 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Data collection Data collection: A village census is often necessary to identify eligible households. Baseline helps collect high-quality contact information to track households in the long term. Collect as many ‘hard’ outcomes as possible (i.e. not just self-reported). Ideas? Development Economics 52 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Haushofer and Shapiro (2016) Main evaluation of Give Directly program. Eligible households are selected based on having a thatched roof. We randomized at both the village and household levels; furthermore, within the treatment group, we randomized recipient gender (wife versus husband), transfer timing (lump-sum transfer versus monthly installments), and transfer magnitude (US$404 PPP versus US$1,525 PPP). ! We focus on the household level randomiztion for this lecture. They measured impacts directly after the end of the transfers (one year after baseline). Development Economics 53 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Main treatment e↵ects ) Strong e↵ects on consumption, assets, revenue, food security, and subjective well-being. No change in Health and education investments. Development Economics 54 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Treatment e↵ects on consumption ) no measurable increase in alcohol or tobacco consumption (measured using randomized list experiment, more to come). Development Economics 55 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Treatment e↵ects on assets ) some investments in productive assets and many new roofs. Development Economics 56 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Treatment e↵ects on assets ) there is no e↵ect on business creation. Development Economics 57 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Take-away E↵ects on a range of economic outcomes is very positive (including consumption). No increase in consumption of ‘temptation’ goods. No change in labor force participation or occupation, increased productivity of existing businesses. The Give Directly program is rated as highly e↵ective by Give Well an organization evaluating development organizations. However, recently they decided that, for the funds Give Well administers, they have more e↵ective interventions. Development Economics 58 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Three year e↵ects (Haushofer and Shapiro, 2018) Economic e↵ects on treated households persist at roughly the same size. They now find evidence of negative spillovers to non-treated eligible households. ) Lower consumption and food security possibly due a reduction in productive assets. Development Economics 59 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Limitations of the study Relatively small sample given the number of di↵erent treatment configurations. No real long term outcomes only (what about poverty traps?) Control group is not cleanly defined. Limited external validity due to the geographic focus. Development Economics 60 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Discussion Complex poverty alleviation programs do seem to be more reliable at overcoming poverty traps, but are much more expensive. If the goal is to mitigate transitory poverty cash transfers seem like a good option (the e↵ects on consumption are much larger). Both types of programs have many have positive side e↵ects (more on this later). Development Economics 61 Extensions Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Other things to learn about cash transfers Intervention parameters (tinkering): Labelling of transfers Conditionality Selection of eligible households Are there general equilibrium e↵ects (i.e. what are the e↵ects if we were to treat a large fraction of individuals). ) Other studies have answered these questions. Development Economics 62 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Extension 1: Long-term results of cash transfer program in Uganda (Blatman et al., 2018) They study the e↵ects of a 400 USD cash transfer to individuals to start skilled trades. They find initially strong e↵ects on income and skilled employment. After nine years, e↵ects on assets and skilled work persist but there are no e↵ects on other outcomes. Development Economics 63 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions The control group catches up ) grants acted more as a kick-start than a lift out of poverty. Development Economics 64 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions How to interpret this theoretically If it were a poverty trap, we would expect to see a decline of the treatment group rather than the control catching up. If it shifted the steady state, we would expect a permanent divergence. ) Maybe individuals converge to their steady state over the early years of their lives as they accumulate assets. Development Economics 65 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Extension 2: Conditional vs unconditional cash transfers Baird et al. (2014) conduct a meta-analysis to understand if conditional cash transfers have better e↵ects on education than unconditional cash transfers. They e↵ectively average treatment e↵ects across all included studies. They use data from 35 di↵erent experiments and quasi-experiments to study the impact on enrolment, attendance, and test scores. There are many more conditional programs (30) than unconditional programs (9). The study is a bit dated, but still useful. Development Economics 66 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Average e↵ects on attendance ) E↵ects of CCTs are larger but the di↵erence is not significant. ) CCTs are more e↵ective if conditions are enforced. Development Economics 67 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions E↵ects on test scores ) Learning gains are at most modest and do not significantly di↵er across CCT and UCT programs. Development Economics 68 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Extension 3: General equilibrium e↵ects of cash transfers One worry about cash transfers is that they could lead to an increase in prices if rolled out at scale. ) This would lead to negative spillovers to non recipients and diminish the value for recipients. Give directly implemented a large scale experiment in rural Kenya to test this hypothesis. The individual treatment is a one-time 1000 USD unconditional cash transfer. Development Economics 69 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Randomized saturation design Note that they now treat all eligible individuals for ethical reasons. Development Economics 70 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Randomized saturation design Randomized saturation designs allow the study of general equilibrium e↵ects. They randomly vary the proportion of treated individuals within a given unit. ) The allows to identify the e↵ect of treating a larger proportion of the population. Here, their main specification uses the total amount of cash dispersed within 2km of the household as treatment (instrumented by treatment assignments). Development Economics 71 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Spatial distribution of villages Development Economics 72 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions E↵ects on prizes Prices rise but only for non-tradables and in places without good market access. Development Economics 73 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions E↵ects on economic HH outcomes ) Spillover households increase their expenditure, assets, and income substantially. Development Economics 74 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions E↵ects on non-economic HH outcomes ) There are no negative e↵ects on spillover households. Development Economics 75 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions Overall e↵ect on the economy The transfers were worth about 15% of local GDP (hence very large). They estimate that the fiscal multiplier is 2.7. ) The economy grew by 2.7 times the transfer amount. Prize inflation was moderate and did not o↵set these gains. ! Implementing cash transfers at scale might have additional positive e↵ects. Development Economics 76 Poverty Trends Living in Poverty Poverty traps Modern poverty alleviation programs Cash transfers Extensions What did we learn today? One-o↵ poverty alleviation programs can but do not always have long-term e↵ects. ! The interventions are relatively complex and expensive to implement (though still worth it). Cash transfers are an efficient if often temporary way of alleviating poverty. Scaling such interventions might have additional positive e↵ects, though the evidence is not uniformly positive. Development Economics 77