Development Economics Lecture Notes Review

Flashcards on development, poverty, growth theory, inequality, foreign aid, impact evaluation, education, labor markets, private sector development, microcredit, gender and ethnicity, and informal institutions.

Characteristics of the Global Poor

The majority of the global poor live in rural areas, are poorly educated, employed in agriculture, and are under 18 years of age.

Challenges in Measuring Poverty

Challenges include poverty being multi-dimensional and often measured at the household level, ignoring within-household distribution.

Consumption Patterns of Poor Households

Poor households spend a large share of consumption (50%-78%) on food, but are not necessarily maximizing calories.

Possible Reasons for Little Specialization Among the Poor

Risk-spreading between sectors and lack of access to capital to scale up.

Reasons for High Entrepreneurship Among the Poor

With few skills, it may be easier to be an entrepreneur than find a job.

Reasons for Low Investment in Education Among the Poor

Illiterate parents may not realize their children aren't learning much.

Reasons for Low Savings Among the Poor

Saving at home is difficult (robbery, family use), and there are self-control problems with temptation goods.

Reasons for Limited Migration Among the Poor

They may remain close to social networks for informal insurance, or face loneliness and poor temporary conditions when migrating.

Potential Policies for Poverty Reduction

Food subsidies, education, health, credit, land reform, or infrastructure (water, roads).

Key Equation for Capital Stock Changes

Capital stock changes due to investment and depreciation.

Harrod-Domar Growth Model

Defines savings rate (s). In this model, the growth rate (g) is approximately (s/θ) - n - δ, where θ is the capital-output ratio and n is the population growth rate. s, n, and δ determine the growth rate.

Solow Model

In the long run, the growth rate (g) is 0. s, n, and δ only affect the steady state level of income, not the growth rate. A limitation is that this doesn't align with observed growth in rich countries.

Solow Model with Technological Progress

Accumulation of capital alone isn't enough for sustained long-run growth due to decreasing returns. Technological progress (π) is introduced, affecting the efficiency of labor.

Solow Model with Human Capital

A third factor of production, human capital (H), can be added. This changes the production function and accumulation equations.

Mankiw, Romer, and Weil (1992) Findings

Differences in saving rates, human capital investment rates, and population growth can explain nearly 80% of the variation in GDP per capita.

Caselli (2005) Suggestion

About 2/3 of variation remains unexplained when using a broader measure of human capital, attributing this unexplained portion to Total Factor Productivity (TFP).

Conditional Convergence

Convergence when controlling for differences in parameters like saving rates, human capital, and population growth.

Limitations of the Solow Model

Not fully explaining huge income differences across countries and assuming differences in parameters (s, n, δ, π) without explaining them.

Lorenz Curve

Plots cumulative population percentage (sorted by income) against cumulative income percentage.

Mean Absolute Deviation

Uses distance from the mean income; fails the Pigou-Dalton principle for transfers above the mean.

Coefficient of Variation

Uses standard deviation relative to the mean; satisfies all four principles and is Lorenz-consistent.

Gini Coefficient

Based on differences between all pairs of incomes; satisfies all four principles and is Lorenz-consistent.

Impact of Inequality

Inequality can lead to inefficiency by preventing low-wealth/income individuals from fully exploiting their capabilities.

Banerjee, Duflo, and Kremer's Contribution

Using the micro experimental approach for global poverty alleviation.

Burnside and Dollar (2000) Conclusion

Aid has a positive impact on growth in countries with good fiscal, monetary, and trade policies but no effect with bad policies.

Robustness of Burnside and Dollar Findings

The findings are not robust to modifications of the sample, definition of aid, definition of good policy, or time periods.

Reasons for Focusing on Aid Amount Disbursed

Direct beneficiaries (the poor) have little voice to report on productive capacity improvements, and donor countries may have objectives other than growth.

Suggestion Regarding Aid Distribution

Aid should be given more selectively where it can work.

Ways to Improve Aid Effectiveness

Involves imposing conditions (conditionality) or evaluating effects after completion (evaluation).

Impact Evaluation Definition

Impact evaluation is presented as a way to break down the problem of global poverty into smaller, testable questions.

Measuring Impact

Evaluating impact determines whether intended changes in outcomes are achieved, moving beyond just measuring inputs and outputs.

Aim of Impact Evaluation

Establishing causality and attribution – the changes in outcomes directly attributable to the program.

Counterfactual Definition

What outcomes would have been without the program.

Retrospective IE

Looks for treatment and comparison groups after program implementation. Limited options for counterfactual identification.

Prospective IE

Designed before program implementation, allowing for better counterfactual identification methods.

Efficacy Studies

Conducted in specific, controlled settings to test proof of concept or key causal mechanisms. Small scale, limited generalizability if scaled up.

Effectiveness Studies

Programs evaluated under normal circumstances using regular channels, aiming for findings generalizable to a large population.

Theory of Change

Outlines how the project should achieve its goals via a results chain (inputs -> activities -> outputs -> outcomes -> final outcomes).

Inputs

Resources (money, staff).

Activities

What the program does (training, distributing goods).

Outputs

Direct products delivered to beneficiaries (vaccines, training hours).

Outcomes

Changes in behavior or conditions after receiving outputs, not fully controlled by implementer.

Final Outcomes

Ultimate goals, influenced by multiple factors.

Specifying Evaluation Questions

Traditionally, impact of the full program on final outcomes vs comparison group. Mechanism experiments test single causal links within the theory of change.

S.M.A.R.T. Indicators

Specific, Measurable, Attributable, Realistic, Targeted.

Requirements for an Instrument (z)

Correlation with the variable of interest x (relevance) and uncorrelated with other determinants of the outcome y (exogeneity/exclusion restriction).

Randomization in RCTs

Randomization between potential participants ensures that, on average, the potential outcomes of participants are the same as non-participants.

Returns to Education

Involves measuring the causal effect on earnings, health, fertility, beliefs, etc.

Improving Education

Policies to increase access and enrolment rates (quantity) and how to deliver quality education to the poor (quality).

Intuition Behind Natural Experiments

Use an instrument for education to identify its causal impact on income.

Requirements for Natural Experiments

Well-defined, explicit rules for inclusion in the treatment. Selection into treatment should be by an outside factor, not individual manipulation.

Empirical Strategy (Simple Differencing)

Compares average outcomes before and after a policy change.

Duflo (2001) Identification Strategy

Uses two ingredients: date-of-birth and region-of-birth.

Empirical Strategy (Difference-in-Differences)

Compares the change in outcomes for the program's target group in high program-intensity regions to the change for the same cohort in low-intensity regions

Potential Reasons for Low Attendance Rates

Credit constraints, bad infrastructure/transportation costs, opportunity cost, informational constraints, low quality of education.

Jensen (2010) Results

People underestimate returns to education (perceived < actual).

Jensen's Experiment Conclusion

Students from richer families increased their schooling when provided information about actual returns. Students from low-income families did not increase their schooling.

Quality of Schooling

Teacher absence rates are high in developing countries.

Why Don't Inputs Work (On Average)?

Inputs like textbooks or flipcharts may not help students at the bottom, who have more basic needs. E.g., in Glewwe et al., textbooks were in English, and some students couldn't read.

Remedying Education

Accounting for Student Differences

Balsakhi Program

Targets inputs to weaker students.

Teacher Absenteeism

Temporary para-teachers (annual contracts) in NGO schools might allow for greater oversight and incentives as they don't form an entrenched constituency.

Summary on Education

High returns to education, demand for education might be lower than it should be, low-cost interventions can improve quality, teacher incentives are important.

Characteristics of Rural Labor Markets

Tasks require physical strength, involuntary unemployment, not all work is monitorable, two contract types coexist, substantial uncertainty and seasonality.

Capacity Curve

Shows the relationship between nutrition and capacity to perform productive work.

Main Idea Linking Nutrition and Work

Nutrition is affected by income, which in turn affects earning capacity.

Subramanian and Deaton (1996) Finding

Poor households in India who are 10% richer spend 7% more on food.

Impact of Increased Income on Calorie Consumption

Even among very poor people, increased economic well-being has a positive but not huge impact on calorie consumption; it's not one-to-one.

BRAC Program Definition

Targets the poorest women in villages, providing assets, training, and other support.

BRAC Program Components

Asset transfer, asset-specific training, free healthcare, training on rights/laws/politics, free legal aid, and village committees.

Graduation Definition

They are left on their own.

BRAC Graduation Program

Targets the poorest women in villages, providing assets, training, and other support to improve ultra-poor women's economic and social situation multi-dimensionally and break poverty traps..

Poor Women's Job Opportunities

Poor women work fewer days due to the seasonality of casual jobs.

Take-Aways about Graduation Programs

Assets, training, and support to the ultra-poor can help them move out of poverty.

Blattman and Dercon (2018)

Examine impacts of industrial and entrepreneurial work based on experimental evidence from Ethiopia.

Exporting/foreign-owned firms pay higher wages and that factory jobs may be preferred over alternatives

Verhoogen, 2008, Atkin, 2009

Private Sector Development/Entrepreneurship

High rates of self-employment exist in developing countries.

Findings About Capital for Private sectors

Firms were severely credit constrained, and the marginal rate of return to capital was very high for them.

Impact of Capital Stock on Profit

Treatment increases real monthly business profit by around 6% on average. This is greater than monthly interest rates observed in urban areas.

McKenzie & Woodruff (2017) Results

Business practices are highly correlated with firm performance.

improve practices and returns to capital

Business training

Financial Markets or Microcredit

Credit markets link savers and investors.

Credit is

needed for fixed capital, working capital, and consumption credit

Vegetable Vendors in Chennai (Karlan and Mullainathan, 2012)

69.4% of the sample takes daily loans, 65.7% take daily loans for >15 days/month.

Importance of Small Loans

This example highlights a potential debt trap where high interest makes repayment barely possible..

Credit in Lusaka

Most never applied for a loan. Of those who did, only 1/5 went to a bank, even rarer during Covid.

High Rural Interest Rate

Thailand (Siamwalla et al.) informal rates min 25% vs formal 12-14%, India (Dasgupta, 1989) average moneylender rate 52% vs max deposit 24%.

Borrower Treatment between Rich and Poor

Poor get smaller loans (rationing) and pay higher rates (exclusivity). India (Timberg & Aiyar)

What's Missing about Loans

Imperfect enforcement.Limited liability.Asymmetric information

Limited Liability

Borrowers' downside, artificially encouraging too much risk and is related to strategic default

Adverse Selection

If lenders can't distinguish risky from safe borrowers, they set a single interest rate based on the average risk. This high rate drives out safe borrowers, leaving only risky ones, increasing default risk further

Ex-Ante Moral Hazard

Borrower takes unobservable action before investment affecting success.

Experiment on Asymmetric Information (South Africa)

Major South African lender sends direct mail offers to 58,000 former clients with good repayment histories.

Canonical Microfinance Model Features

1. Leans almost only to women. 2. Group lending with joint liability. 3. Weekly repayment schedule. 4. Regular meetings where members forge social bonds. 5. Very small loans initially, increasing over time. 6. Extensive monitoring by credit officers, with incentives based on clients and repayment rates. 7. High interest rates (at least 20% a year).

Potential Effects of Group Lending

Screening (women select reliable members) (Ghatak, 1999), Peer- monitoring (members monitor each other for free).

Feigenberg, Field and Pande (2012)

100 microfinance groups in India randomized into monthly repayment/meeting or weekly repayment/meeting. Weekly meetings (vs. monthly) reduce default rates

Microfinance requirement (MFIs)

Savings

When legal, MFIs require savings, which

Helps to build collateral.

Banerjee et al (2015): Evaluation Design

Microcredit randomly introduced in 104 slums in Hyderabad, India (2005).

Baseline Characteristics

After 18 months, other MFIs were present or starting in control slums. Treatment and control slums were very similar at baseline.