lOMoARcPSD|4405180

Summary Development Economics: lectures

Development Economics (Vrije Universiteit Amsterdam)

StuDocu is not sponsored or endorsed by any college or university

Downloaded by Gisha Bekele (gishabekele1@gmail.com)
 lOMoARcPSD|4405180

HC 1. Poverty http://web.uvic.ca/~kumara/econ420/chap8.pdf

Wat is poverty?
- Below the poverty line (= the level of minimum income) = earn less then 1 dollar a day
- Basic needs: food, shelter, health
- No future prospects for growth
- Poverty vs. Vulnerability
- Civil Rights

“Voices of the poor?”

- List of characteristics of poverty (given by ‘poor people’)
- Poverty is multi-dimensional (this makes poverty comparisons difficult, you need a single
indicator for a complete ordering)
- Economists: use consumption (hard to measure income for the poor; no steady income
flow and informal work).
1. Income difficult to measure for certain jobs
2. Consumption covers self-produced goods
3. Income fluctuates more than consumption and therefore consumption is a better
indicator of welfare.

Issues: No direct relation between income and nutrition

- Amartya Sen: look at capabilities of people: What matters is not the things a person has,
but what a person is or can be and does or can do.
Functioning’s: what a person does or can do with commodities that they own or control.
Looking at real income or even the levels of consumption of specific commodities cannot
suffice as a measure of well-being. These are of little value if they are not what consumers
desire. Capabilities: the freedom that a person has in terms of the choice of functioning’s,
given his personal features and his command over commodities.

Cut-off point: poverty line. Poverty measured is based on a poverty line.

- Absolute vs relative
• absolute poverty line is a threshold, usually expressed in terms of income that is
sufficient for basic needs, that is fixed over time in real terms. In other words,
it’s adjusted for inflation only and doesn’t move with economic growth, average
income, changes in living standards or needs.
• A relative poverty line varies with income growth or economic growth. Example:
Income equal to half the mean income in the population.
- Objective vs subjective (subjective = ask people when they feel poor).
- Temporary or chronic
- Household vs individual

- Measurement:
1. 20th percentile of income distribution
2. Costs of minimum food basket
Food energy intake method : 2100 calories a day is the norm. (Absoluut)
Calculate calorie consumption per capita as a function of household income per capita.
Plot food energy intake against consumption expenditure to work out the level of
consumption necessary to achieve the minimum calorie intake
The problem with this measure: it will alter with differences in tastes, activity levels,
relative prices, migration

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

• Urban poverty line: more consumption expenditure for same calories (compared to rural
poverty line).
Urban consumers tend to consume less calories per day at each income level and this
can typically only be partly explained by higher urban prices. Hence, the urban
poverty line tends to be high relative to that in the rural area. Reflecting possible
differences in preferences (taste differences) and need for calories across urban and
rural areas).
3. Decreed (=vastgesteld) minimum wage
4. Ask people what is minimum necessary income

Differences in poverty lines may not equal differences in the costs of acquiring the same
basket of basic foods

Why is there a difference between urban/rural?

- Production in rural area; cheaper
- Different foodprices: food is more expensive in city
- More choice in urban areas and more processed food.
- Taste differences!

Problem with using a poverty line for comparisons:

- Cost of living differ considerably in poor countries
- Commodity bundles consumed differ considerably across countries
- Cannot judge whether growth actually affects poverty or not
- Poverty is measured using survey data while growth comes from National Accounts

Non-food consumption: lots of choices, murky standards

Thus: adjust food PL based on observed non-food consumption (see graphs).

Richer people – and richer countries – tend to have higher poverty lines
- Amongst poor countries, the poverty line does not change much with income - absolute
consumption needs dominate.
- But as incomes rise poverty lines start to rise as well.

Measures of poverty
- Headcount: absolute number of people under the poverty line
- Headcount ratio: percentage of a population under the poverty line
• It does not tell us how far one is below the poverty line does not show relative
distribution of the poor (how poor are you??).
• Policy trap: targeting on lower headcount can lead to undesirable results
• Policy biased towards simple HCR
- Poverty Gap Ratio: calculates how much money is needed to eliminate poverty through
transfer
- Foster-Greer-Thorbecke:
• Alva >1, weak transfers principle. Weak transfers principle states that a transfer of income
from any person below poverty line to anyone less poor, while keeping the set of poor
unchanged, must raise poverty.

Poverty GAP Ratio - Issues in alleviating and measuring poverty:

BUT still no special attention for the very poor.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

- Target the people who are actually poor (how do you find them?).
Conclusie: Measures of poverty distributional sensitivity matters

The face of poverty

Household size: correlation with poverty (also for female headed households).
Location: Rural has higher poverty than in cities (urban). In cities informal markets have
high poverty. Poor lack productive assets. Poor lack of human capital.

Poverty Traps
Basic idea; people are below a level where investing in growth is not rewarded.
Reasons why people are stuck;
- no access to collateral, credit market requires collateral.
- entry costs in the labor market (proper clothes, degree, literacy etc.)
- increasing returns not captured (returns from education only start after 4 years of
education, not before that).
There is some kind of threshold level that the poor lack.
Poverty trap: Poverty →under-nutrition → lowers the ability to work and earn an
income →under-nutrition

Capacity curve Ray

- Relates income and work capacity
(productivity)
- Higher income better nutrition
- Better nutrition: first used by the body for
basic metabolism. Then only it translate in
higher capacity.
- As a result, the work capacity is convex, and it
intersects the 45degree line from below.
- Middle of the curve (unstable equilibrium),
if you drop below middle point, your productivity will move towards 0. If you earn a little
bit more, you will move towards the highest point.
- The 'vicious cycle of poverty' is: Lack of labour market opportunities makes for low
wages. Low wages lead to low work capacity. But low work capacity in turn closes the
access to employment

Intra-household allocation
- Lifeboat problem: can hold only two people and there are three individuals to save. One
person must die. The capacity curve gives us a clear idea of how the nutrition problem to
promote unequal allocations. Within the household people have to make tough decisions
about who to allocate each resource to and what yields the highest productivity.
- Least productive may suffer: children, girls, old, ill

Dotted line on the curve (slide 38); equal division

(point D), can also allocate it to one person (y),
you end at point E. Divide it between two people
(y/x), you end at A. Total return (productive work
capacity) is highest when you only give it to one
person.
UP to a certain point B! Then it is better to use
equal division. ???????

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Intergenerational cycle of poverty

- poverty malnutrition of young children
- Alderman, Hoddinott and Kinsey (2006): malnutrition at a young age lower growth,
poor health, less cognitive development less schooling (less motivation, energy, skills);
exacerbated by immediate effects of poverty and malnutrition by immediate effects of
poverty and malnutrition
- less schooling lower future income-generating capacity lower future income
poverty as an adult, etc.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 2. Inequality
Veil of ignorance – John Rawls
You could be rich or poor; want an equal situation where on average you will be best off!
Maximin; maximize output for the worst off individual (extreme risk aversion).

• Intertemporal issues: changes over time

Used to be that US was more mobile than EU but not anymore.
In general; studies find that mobility is no longer higher in the US than in Europe.
But the perception is still that US is more mobile; therefore people are more tolerant towards
inequality (since they think they can move out!).

What is inequality?
Inequality is a function of an income distribution (can be any kind; wealth, health, income
etc.). I = I(y1,y2,..,yn)
Convert distribution into an inequality measure by; agreeing on acceptable or desirable
properties of inequality measure in order to compare inequality (between countries and over
time).

Inquality measures: 4 axioms:

- Anonymity principle: independent of who
- Population principle: proportion not absolute
- Relative income principle: absolute income levels have no meaning
- Dalton principle of transfers: A regressive transfer of income from a poorer to a richer
person and result in a more unequal distribution, should increase poverty.

Some examples
- Range; take two extremes and normalize it with mean income (only satisfies 1 and 2).
- Kuznets ratio; poorstest 20%, richest 10%
- Mean absolute deviation: mean absolute deviation from the mean, take the average and
normalize with mean income; does not satisfy Dalton
- Coefficient of variation; Dalton does not hold, there are regressive transfers that have 0
impact.
- Theil index
- Gini-coefficient: A/(A+B) is a measure of statistical dispersion. The Gini coefficient
measures the inequality among values
of a frequency distribution (fe,
income).
- A Gini coefficient of zero expresses
perfect equality, where all values are
the same (fe, where everyone has an
exactly equal income). A Gini
coefficient of one (100 on the
percentile scale) expresses maximal
inequality among values (for example
where only one person has all the
income).
- Difference between all pairs of incomes and totals (absolute) differences.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

• Lorenz Curve : a graphical representation of the cumulative distribution function of the

empirical probability distribution of wealth; often used to represent income distribution.
• Lorenz criterion: If a Lorenz curve of a distribution lies at every point to the right of
another Lorenz curve, the former distribution is more unequal.
An inequality measure is consistent with the Lorenz criterion if and only if it
satisfies 4 principles: Anonymity, Population, Relative Income, Dalton
- The Lorenz criterion satisfies these principles because it looks at the variable of interest
(anonymity) and percentages (population and relative income). Compare two
distributions, where one of the distributions is created from the other distribution through
one regressive transfer. The Lorenz curve with a regressive transfer then lies to the right
of the other Lorenz curve. From the Lorenz criterion we then know that inequality has
increased (Dalton).
- If Lorenz curves of two distributions do cross, then Lorenz criterion cannot be applied. In
this case different inequality measures which still satisfy all four criteria may give
contradictory results.

The effect of income on inequality: Kuznets

- Lewis 2 sector model:
Dual economy: labor surplus and subsistence wages in rural sector, modern urban sector
with higher wages and profits.
- Growth in modern sector causes higher inequality in the short-run because rural wages
increase slowly and modern sector is initially relatively small.

Inequality and income growth

3 reasons why inequality may have impact on economic development:
1. Inequality affects aggregate savings if the marginal propensity to save varies across
income
2. Inequality may affect the access to credit if credit markets are imperfect (are require
collateral for instance)
3. Inequality may affect the security of property, social/political stability and the demand
for taxation.

1. Savings
- Classical growth model: higher
savings more investment in
capital higher growth
- Common argument: ‘The rich
have higher propensity to save
and therefore for a given GDP it
is better to have inequality so
that some people will save’
- Issue: are marginal savings rates
in- or decreasing ?
????????
Propensity to save increases with income. Higher savings with inequality. Lower savings in
an equal society. Redistributing wealth may therefore adversely affect the savings rate and
therefore economic growth in poor countries.

2. Credit - Inequality can affect growth if the credit market is imperfect.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

The credit market increases growth by providing credit for investments that increase output
(economic growth).
Asymmetric information in credit market: borrowers may exhibit opportunistic behavior,
refuse to pay back the loan.

- High inequality limited access to credit, few entrepreneurs, less demand for labor, low
wages
- Pareto inefficient: credit market failure limits the number of profitable transactions;
lowers aggregate production and income
- Inequality is replicated: poor cannot accumulate wealth because of low wages

3. Political Economy -Negative effect of equality on growth

- Via security of property, social/political stability and the demand for taxation
- Via political process: median voter determines the outcome of political process. If income
distribution is unequal, the median voter is poor demand for redistribution. Result:
higher capital taxes, distortion lower growth.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 3. Worldwide Inequality

Three concepts of world inequality

- Unweighted international inequality (country GDPs per capita)
- Population-weighted international inequality (based on people in country and then look at
relative size) does not take into account the difference within the country, just looks at
average.
- Inequality across all individuals

Classical theory on convergence:

- HO theory; factor and goods price equalization
- Growth models (Solow), capital accumulation
- Technology catching up models
Since 1970s less confidence in convergence
- Growth failures of poorest countries
- Further regional rather than global integration
- Evidence of global divergence rather than convergence
- New theories of divergence based on transaction costs, externalities, indivisibilities,
imperfect markets.

Theory; if there is some kind of lower bound (say 250 as absolute minimum), then looking at
GDP over time (table of pridget); big time divergence.

Attempts at calculating world income distribution

- Previous attempts
- Approximate distribution from first two moments and/or Gini coefficient
- Approximate distribution by matching countries
- Milanovic approach; nationally representative household surveys.

Within-verus between-country inequality

Relative income principle is satisfied in the Gini so the within country inequality should be
the same for international dollars and US dollars. The explanation is; that the Gini cannot be
decomposed completely.

What drives changes in world inequality?

- Gap between rural India and rural China and large rich countries and large rich countries
- Gap between rural India and rural China and urban China

Gap between method 2 and 3 = within country inequality (very important in 1820, less so post
1940s)
In 1820; position of society was most important (not the country). Today; ask about passport.
In the 50s this was at a close point. Post 1950s; again an increase in within country
differences more important.

Based on income there is an increase but not on other measures; life expectancy, television
ownership, education etc. all has gone down.
What is missing? Wealth inequality! Hard to measure.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 4. Rural and Urban

Transitions along with economic development:
- Transformation from agriculture to industry and services with economic development.
- Rapid urbanization around the world (more industrialized countries,also within countries).
- Large rural-urban migration flows; floating population is very big (especially in China).
- Migrant labor in the industrialization process
- Large informal sectors and slums in LDC urban areas

Rural-urban interaction: Lewis (complete story about movement of people and food to cities)
Migration: Harris-Todaro (individual and family decisions about moving).

Lewis model: Based on the duality of the traditional and the modern sector.
Modern: industry new (capital intensive) profit maximiation.
Traditional: agricultural old (labor-intensive) income sharing within the family
(consumption smoothing).
basic concept: transform from low-productivity (rural) to high-productivity (urban) society.

Traditional economy:
Low MPL in agriculture; pulling out labor out of agriculture has negligible social cost. Wages
are low (not equal to MPL). Income sharing within the family. = excess supply of labor.

Capatalist sector:
Industrial expansion can be fuelled by the surplus of labor due to low wages from rural areas.
Start with surplus labor, more to disguised unemployment and then to commercialization

- First turning point; surplus-disguised unemployment: wages will start to rise beyond
minimum.
- Second turning point: into commercialization; industrial sector will start to compete with
agriculture.
- Tension; income of the farmers and food prices and the industrialists. Still need cheap
food to ensure cheap labor to build industry. Need to keep care of your agricultural sector.
- Urban bias; higher taxation on farmers than on urban citizens. Priority on industrial
development; cannot be done without proper food production! Some say; cannot be done
without low wages in agriculture.

Lewis model: Policy Isseus

Agricultural taxation: keeps agricultural wages down and labor supply elastic
Hard to do; hard to measure, not a formalized market, based on what? What is the income?
How to collect; when to get it.
Bad for future agricultural surplus (reduces farmer incentive to invest in agriculture)

Agricultural pricing policy:

Higher output prices, Lower input prices, overvalued exchange rate/export tax
All makes sure that agricultural products stay in domestic market and keep food prices low.

Agricultural reform to improve incentives

(Household responsibility system in china; quotas of what to produce and how much to sell it
for).

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

What is the most important price in the world? Wages of Chinese workers
Discussion on world inflation rate (relatively low for a long time, along with high growth).
In China it took 20 years to move labourers out of agriculture but now wages are starting to
rise. Response in China; re-investment in agriculture, demographic transitions.

In general
- Ensure investment in skills of labourers
- Grow beyond cheap labour industry into more skilled industries and service

Do you think agricultural development is necessary for industrial development?

First ensure proper efficiency in agricultural development, create surplus and then use your
additional labor to invest in industry.

Lewis vs Harris-Todaro
Harris-Todaro addresses urban under-employment and low-wage informal sector.
- 3 sectors:
1. Formal urban sector,
2. Informal urban sector (with unemployment),
3. Agricultural sector (informal too).
In H-T model there will be (i) lucky migrants finding formal sector jobs; and (ii) unlucky
migrants defaulting to informal sector jobs.
- Migration: rational individual (HH) decision under uncertainty: choice between sure-but-
low agr wage and ‘lottery’ - chance of getting a high urban salary. High uncertainty in
moving to the city; if you get a formal job or end up in the informal sector.

- Pull factors; higher wages, better facilities (difficulty of finding a job in town -)
- Push factors; lack of employment in rural areas.

Model predicts: how much migration and employment levels in both formal and informal
urban sectors.

See graph; agricultural and formal wage > modern sector in city; above clearing wage.

Why high wages in the formal city?

- Unions; more organized and concentrated laborers (or insider power), costly to fire
- Minimum wages more enforced
- Investment in training is compensated with higher wages
- Over-pay certain jobs to ensure the right people, talented and motivated

HT - Equilibrium:

Assumptions: no costs of migration and the individual is risk neutral.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Paradox: formal sector job creation may increase urban unemployment if ‘migration
effect’ dominates the ‘soak-up effect’.

Fewer people in the agricultural sector; increase in wage in agriculture; then there will be an
increase in formal sector jobs based on the equations (slide 28).
But the informal sector jobs also increases to keep finding a formal sector job constant.
The increase in formal sector jobs might be offset by a similar increase in informal sector
jobs.

Is there really a paradox?

No clear empirical evidence.
Because; simplified model; people might simply return when they do not find a urban job.
There might not be over-payment of urban jobs after supply increases.
Appropriate analysis but need to take other things into account as well.
Other reasons for migration

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 5. Land markets
Importance of land markets
(a) Inequality of land market and lack of access to land may create strong migration pressures
with consequent political, environmental and economic consequences
(b) Functioning of credit market depends on functioning of land market
(c)Given optimal factor proportions given technology, eg manhour / m2 land, land markets
(and labor markets) can improve efficiency of agriculture
High population density puts limits on inequality. Problems with unequal land holdings
(LA, Asia) and ambiguous ownership and command use (Africa).

Evidence: cultivation types

- Owner-cultivation (no contract): high in Asia, Latin America, low in Africa
- Tenancy (rental contract): regional variation, fixed rent in LA, sharecropping in Asia
• Fixed rent: farmers bear (dragen) the risks of cultivation (teelt)
• Share-cropping: is a system of agriculture in which a landowner allows a tenant to
use the land in return for a share of the crops (gewassen) produced on the land.
- Squatting or tribal or communal tenure

Land tenure contracts Sharecroppoing inefficient?

- General contract form: R = αY + F Adam Smith, Alfred Marshall: sharecropping
Fixed rent: α=0,F>0 inefficient: due to inappropriate provision of
Sharecropping: 0<α<1,F=0 incentives
Pure wage contract: α = 1 , F < 0 - Basic argument: sharecropping is like a tax on
output (labour)
• tenant (huurder) does not receive entire
marginal product
• gives disincentive to work

Case 1: ‘Owner cultivation’

- No rental contract
- Return = f(L); Costs C(L) = c*L
- Profit (net return) = f(L) - C(L)
- Marginal profit = f’(L) - C’(L) =
f’(L) - c
- Farmer maximizes profit: choose L
so that f’(L) - c = 0 => L1

Case 2: Fixed rent contract

- Fixed rent = F
- Profit (net return) = f(L) - C(L) – F
- Marginal profit = f’(L) - C’(L) =
f’(L) – c
- Farmer maximizes profit: choose L
so that f’(L) - c = 0 => L2 = L1
- Conclusion: incentives unchanged
• tenant makes the same choice
as owner cultivator

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Case 3: Sharecropping
- Tenant pays share s to owner
- Profit = (1 - s) f(L) - c * L
- Now the incentives have changed
‘at the margin’: tenant will choose
L such that (1 - s) f’(L) - C’(L) =
(1 - s) f’(L) - c=0
- This time tenant chooses L3 < L1
- Output ‘tax’ discourages labour
effort

Pareto inefficiency of sharecropping

Under sharecropping L=L3
- where (1-s)f’(L3) = c
- total surplus = IH
- owner gets s*f(L) = IG
Set F = IG in fixed rent contract reduce labor input > inefficient because MC≠MR
- owner indifferent
- now L = L2 = L1 with f’(L2) = c
- total surplus = CD > IH
- tenant should prefer fixed rent

Conclusion on contracts
- Sharecropping is not Pareto efficient: by moving to a fixed rent contract it is possible to
make the tenant better off while the owner is indifferent between the two contracts
- Why do we observe sharecropping in practice? Answer: risk aversion

Risk
Many economic decisions involve uncertainty, uncertainty can be expressed in probabilities.
- Expected value of the investment: kans 1 * waarde 1 + kans 2 * waarde 2 = minimum
waarde your prepared to sell your investment. (< 6000 euro)
- Most people are risk averse: the behaviour of indivuduals while exposed to uncertainty to
attempt to reduce that uncertainty.
- Insurance: expected value = certainty equivalent + risk premium

Risk Aversion Risky Projects

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Assume: Landlord: risk-neurtral, tenant = risk-averse.

Risk-aversion: people care about their utility not expected value. Te spread of the risk
matters.
Design share contract that leaves owner indifferent but farmer better of than under fixed R.

Applying risk aversion to tenure-

- Fixed rent R
• p: profit = G - R; 1 - p: profit = B - R
• farmer: Exp return = p(G-R) + (1-p)(B-R) = pG + (1-p)B - R
• landlord: R
- Sharecropping:
• Exp return landlord = s[pG + (1-p)B] = R
• Exp return farmer = pG + (1-p)B - R = pG + (1-p)B - s[pG + (1-p)B] = (1-s) [pG + (1-
p)B]

In good state farmer gets:

- Fixed rent: G-R
- Sharecropping: (1-s)G
- Prove: (1-s)G < G-R

Sharecropping = risk sharing

- ‘spread’ (production risk) of returns is reduce risk sharing with owner
- Farmer is risk averse, prefers sharecropping in this model.
- Landlord can write contract that leaves farmer and himself better off than under fixed rent

Sharecropping could be seen as a second-best solution in view of missing markets.

Crop sharing reduces risks which is beneficial with imperfect insurance/credit market
(risksharing!)
Share cropping increases return to the tenant under bad state, and lowers under good
state.

Conclusion:
Risk can be lowered even further: pay fixed wage to farmer hire the tenant as a laborer
- However:
(1) Landlord may also be risk averse
(2) Incentive problem: the contract must also give incentives. Hired labor takes away the
incentive to work hard. Wage labour must be supervised which is costly.
(3) principle agent problem!
- Contract trades off insurance versus incentives

• Double incentives problem: when we take into account the landlord incentive to
maintain the land so it can be leased out in the future
• Limited liability: when the harvest fails the tenant might not be abloe to pay leads
to risky behavior/ over-investment because debt will have to be forgiven.
→ Credit contracts: raise in good times, lower in bad.
• Screening: menne of contracts in an attempt to select the highly productive tenants
ask higher rent

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

• Renewal and evictions (vernieuwing en uitzettingen): additional tool for landlord.

Can help provide additional incentive for payment to hard work and can protect the
landlord against hidden info.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 6. Land Ownership
Scale and productivity; inverse relationship; Ray conclusion = small scale is more
productive. Very recent; different trend; land-grabbing (produce on very large scale).
This is due to a trend of labelling and information that is required of products that is easier for
larger traders.

Farm Productivity – Scale and productivity: 4 factors

- Technology (economies of scale): Total factor productivity and efficiency
Efficiency; more efficient is closer to the production frontier (tech reasons).
Investments increases efficiency need a minimum scale to increase relative output
to costs. This would imply large scale has an advantage, however, there are also many
products that are scale neutral (coffe, tea).
A second aspect is the difference between market prices: observed wages, not equal to
social cost of wages. Wages are often higher than the social costs (due to organization)

- Imperfect insurance markets: Risk

Fewer incentives will be offered to hired labor relative to owner-cultivation if labor is
risk averse.
Owner-cultivation; more popular in this case when labor is risk averse (hired labor
needs additional insurance against risk that might not exist). Counter argument; owner
cultivator might also try to self-insure by choosing low-risk crops.

- Information (monitoring) problems: Labor

Supervision is hard on larger scale, easier within a family farm productivity of
large scale production with hired labor is lower than on small family farms.

- Nonclearing labor markets: Labor

Opportunity cost of family labor is lower than market price. Small farms have higher
productivity:, small farms employ more labor/m2
Marginal productivity of labor is not the same everywhere; frictions in putting people
in the most efficient spot.

For pure owners there is a clear negative relationship between yields and scale, but not for the
other forms of tendancy. Small farms may have higher yields because less problems with
moral hazard (work effort) and imperfect labor markets (surplus labor).

Counterargument: Pooling land

Small farms could pool land to get benefits of technology, right?
Depends on the source of scale economies:
- marketing is easy; sell products together, share fixed costs equally among farmers, easily
supervised.
- However, not so much for production (some labor sharing, building some things during
low-agriculture times or common infrastructure). Pure farming not that much.
Effort not easily observed, incentive to free ride (people work less hard). Better to work
on your own plot.

Empirical evidence
Look at output per acre (not profits or efficiency); because it is an easy way (production
functions are hard to compare).
Good indicator of how people are able to incentivise their workers and who is better at it.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

- Highest; owner-cultivated, family labor

- Lowest: sharecropping (good on risk sharing but low output).
- In between: large mechanized farms and hired labor.
At a very small scale there could be diseconomies of scale.

Small farms and the future

Farming is still dominated by family owned businesses
But developments may affect this:
- Technology; might take away the need for labor and thus incentive problems.
New seeds; no weeding necessary; only need to plant and not much more labor after that
(good for large farms!).
- Need for certification; easier for larger scale.
- Global integration; need international connections to be competitive (local market cannot
compete). Demand for streamlined products and controlled produce on massive scale
(imposition of standards).
- Weakening of public R&D: locally adapted agricultural techniques are less prominent.
Private investments are very high but are monopoly owned and hard to compete with.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 7. Child Labor

ILO convention: 5-11 years any hours worked beyond one hour counts as child labor. ‘
12-14; more than 14 hours counts as child-labor if it is not hazardous.
Children can work and go to school; give up leisure time.
Worst effect; give up human capital creation during vital part of their childhood.

Most child labor is in agriculture and household chores.

Girls work more than boys overall, but also more in unpaid labor.

Child labor and poverty: Child labor has a strong correlation between income of
household, child labor is a manifestation of poverty.
Downward trend; growth of Asia has a large reduction of poverty and also child labor.
Modeling can: help to understand household decision making, efficiency issues, costs and
benefits of policy intervention, serve as a guideline for empirical analysis.

- Human capital model: Invest now to have higher outputs in the future (see graph)
- Efficiency issues: Maximization of societal welfare and output = social welfare
economics (function of all utilities).
- Poor society; costs might be higher and returns lower because there is no access to jobs
that provide output.
- Government can solve credit constraint issues; this will be beneficial due to public
benefits.

Economic Problem:
- Concern (Udry, 2003): child labor sacrifices future welfare of the child in exchange for
current benefit for household.

- Dynastic trap (vicious cycle)

Udry paper; not socially efficient to employ children.
Arguments:
1. Positive externalities to education
2. Allocation of benefits between generations within family
3. Financial markets imperfections.
Parents are maximizing their own income and are not far-sighted enough or predict help from
their children later on; therefore less willing to invest in their child’s education.

Causes and consequences:

Increase wages for adults
- Strong cross sectional variations, but increase agricultural prices can both work in 2 ways
on child labor.
Income and child labor correlation does not imply income is the underlying cause.
- Returns to education might be higher for rich children
- Mayby quality of education is lower for poor
Look at shocks, why?

Policies:
- Increase adult incomes; usually also increases output from child labor; increased
opportunity cost of sending people to school.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Why look at external factors? Correlation does not mean causation; need to isolate specific
factors and their impact on child labor (identify the income effect separately, for example).
- Reduce cost of education
Conditional cash transfers, provide food, public transport
- Ban child labor
Laws against education. Difficult to implement, immediate effect is less opportunities
for poor families.

Multiple equilibria for investment in education - Why might there not be one optimal level?
- Credit constraints
- Costs of education might be different for individuals (distance to school for example)
- Quality of education (No supplies, no proper building, large classes, low teacher quality)
- Anticipated labor demand conditions/ labor market (less demand for labor in school, learn
more from learning the ropes at the farm).
- Information (parents lack fundamental information about costs and benefits; more
uncertainty involved in estimating output from education, the increase in income might be
hard to know and be sure about the investment in education).
- Discount rate (different value of the future depending on life expectancy, for example)
- Social stuff (girls don’t need to go to school, facilities at the schools)

Solutions:
- Scholarships, cheap loans
- Public transport (cheap or free), build more schools
- Improve supplies, teacher quality, budgets for schools
- learn
- Improve access to diverse jobs, infrastructure for jobs
- Improve children about the benefits of
education

Economic models
- Basu and Tzannatos (WBER)
- Parents will only send their child to
work if there income falls below
minimum levels
- adult and child labor are substitutes

Multiple equilibria model: no child labor

N, little child labor Sl, high child labor Sh.
Subsistence consumption level s: below
this level there will be a need for children to
work. Against economic theory; wages go down; people now have to work more (to make up
for the low wage). Now the supply curve is downward sloping.
E2= no child labor. E3 = child labor, still subsistence s is reached.
People cannot change the wages, the government can say; ban child labor, move the economy
back to E1.
Assumption; there will be less labor and you will need to pay people more to attract enough
labor (not sure if this is realistic).
As soon as you pass E3 it will be easy for employers to lower wage; there will be even more
demand, move to E2.. E3 is therefore not a stable equilibrium.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Social norms can also lead to multiple equilibria

If there is a lot of shame in letting your child work; harder to do it. But if everyone sends their
children to work it is not such a bad thing.

Conclusion:
• Banning can be appropriate instrument in general equilibrium model
• Wage effect might be too strong
– Social norms can also lead to multiple equilibria

General conclusions
- Good schooling, free school meals
- Compulsary education: monitoring
- ‘Work-and-school’, poor countries
- Higher adult incomes (decreases child labor
- Ban hazardous work; total ban may work in better-off countries
- Import restrictions are sector specific, may hurt children
- Need a push from outside to make the whole economy change
- Enforce laws for safety, minimum benefits etc

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 8. Credit
Why is credit important?
- Rural poor: low incomes, low savings/assets, income fluctuations/shocks, agricultural
production takes time
Liquidity problems, need money to smooth consumption and production

Demand for credit:

- Fixed capital: land, machinery, buildings
- Working capital: cash needed for financing inputs
- Consumption credit: needed for survival after shocks

Is access to credit really a constraint to development?

If I had more money I was richer and credit being a constraint.
Is it profitable to borrow money to expand your business? Then it is a credit constraint.
There is only a credit constraint if there is actual economic growth due to investments.

Who provides credit?

Formal lenders (banks, commercial banks), informal lenders (moneylenders, family, traders).

Vietnam example
Many loans in the informal sector, less so as the country developed (more formal loans).

Empirical observations
- Sizeable gap between lending rates and deposit rates within the same sub economy
- Extreme variability interest rate
- Production and trade finance are the main reasons giving for borrowing
- Credit limits: rich people borrow more at lower interest rate
- Many people want to borrow at formal interest rate
- Very high share of credit transactions between friends / family, many without explicit
interest
- Low formal interest rates, high moneylender rates

Policy questions
- Should everyone get credit?
- Are high interest rates exploitative or efficient?
- Should government intervene?

Who should get credit?

• Efficiency: cost of capital is the minimum interest rate
– $0.20 per dollar: projects with return of $0.15 excluded
• Uncertainty: even high return projects fail
– Expected return: 75%*0.40 + 25%*0 = 0.30

Credit market imperfections

Monopoly: costs high or because it is a monopoly? (high transaction costs, default,
information costs).
Information (agency) problems: moral hazard, adverse selection (limited liability).
Limited liability: Lending to poor strangers is risky

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Limited liability LL
Borrower prefers risky project because they are not punished for losses but rewarded for
wins.
Banks; lower interest rates to have 50% safe and 50% unsafe borrowers (higher there are only
risky projects left). > example of credit constraints; profitable projects are available but no
one is investing

LL: Lender Strategies

• LL creates excessive downside risk for lender; what to do?
1. Collateral (onderpand): reduce information need
2. Information: reduce information asymmetry (= deals with the study of decisions in
transactions where one party has more or better information than the other, access to
different information.)
– Screening, monitoring
3. Rationing: take into account borrowers’ default (verzuim) incentives when writing
contract

Credit rationing
• ‘At the going rate of interest the borrower would like to borrow more money but is not
permitted to by the lender’ (Ray, 548)
• Result: credit constraints; less growth, more poverty!
• We discuss two types:
– Default related (moral hazard: is a situation where a party will have a tendency to take
risks because the costs that could incur will not be felt by the party taking the risk)
– Borrower risk type (adverse selection It refers to a market process in which undesired
results occur when buyers and sellers have asymmetric information; the "bad" products
or services are more likely to be selected)

Default and enforcement f ( L) − L(1 + i ) ≥ A

L = loan size, A = earning at outside option

No default constraint: f(L)- N/(N-1) L(1+i) ≥ A

N = n periods; after default in period 1 no more loan, A = profit when borrows from other
source

If A goes up: default goes up

(harder to meet default constraint).
N increases: decreases defaults
(number of periods where you get
punished = N).
Default constraint lowers the
amount of credit given out (below
optimal level).

Credit rationing in equilibrium

• Money lender keeps interest
rate at participation maximum
• But reduces loan size to L1 to also meet the pay back restriction
• At this interest rate borrower would want to borrow more: credit rationing in equilibrium

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Screening and enforcement

- Need collective action and information sharing to identify bad borrowers.
- If default information is spread rapidly, bad behaviour is punished screening.

Adverse selection and rationing

- High-risk vs. low-risk borrowers; adverse selection (information asymmetries).
- See model; safe and risky formulas.
- As a result of increasing the interest rate your customers will be more risky.

Zie slides voor extra uitleg.

Group lending
• Micro finance institutions lend to groups
– Groups form by themselves, they know what types others are
– Are liable for each others default
• Group Lending reduces credit rationing as a result of adverse selection
– Assortative matching ensures that risky types and safe types form separate
groups
– Joint liability reduces risk of default for risky borrowers
– Banks can reduce their interest rate, making it more attractive for safe types to
borrow

Solutions to deal with these risks:

- Group lending: monitoring, joint liability
Assertive matching: ensures that risky types and safe types from separate groups
Interest rates can be lower because every member has to pay if one defaults (and are
then excluded from the market if they don’t).

Interlinked credit
• Pay back in output
– Often provided by landlord, trader
– Option in regions where charging interest rates is forbidden
– Allows to keep more renters in the market as risk of strategic default is reduced
• Has the same negative incentive effects on effort, as output is taxed (sharecropping)

Conclusion:
• Lenders face LL and asymmetric information
• Strategies
1. Collateral (land titling !)
2. Information:
• Segmentation, exclusive dealings : repeat lending to fixed clientele (switching
costs are high)
• Interlinked transactions: moneylenders are landlords, traders, shopkeepers
• Result: interest rate variation plus local monopolies based on information
3. Increase interest rates
4. Reduce loan size to prevent strategic default
5. Interlinked credit
6. Group lending

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Empirical validation
Default = function (interest rate charged)
Positive relation could be due to
– Moral Hazard
• Defaulting becomes more attractive relative to paying back loan
• Increase collateral requirements, promote dynamic incentives
– Adverse selection
• More riskier borrowers apply if interest rate is high
• Loan guarantees, advanced screening, subsidize loans
As interest rate increases, chance of default increases (can be due to; moral hazard or adverse
selection).

Consumer credit field experiment

Adverse selection is something that happens before you apply for the loan.
Moral hazard happens after you have applied for the loan.
They split up these two steps on their experiment to isolate which imperfection plays a larger
part in defaults for small loans.
Dynamic incentive had the most effect on default.

Impact of micro credit

• Micro credit programs have been donor’s favorite, but are now often criticized.
– Large demand for loans is evidence of success
– But very little evidence of permanent income gains

Is credit the answer?

• Credit provides resources for investment
– But at typically very high interest rates
– Default can throw people back into poverty trap
• Credit comes with other services
– Business support services (group work, technical)
– Commitment device (have to plan ahead)
• Are it the other services what makes credit attractive?
– Savings commitment devices

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 9. Risk and Insurance

Types of risks; climatic, accidents, crime, economic (unemployment, price fluctuations etc.),
health, social and political (civil unrest etc.).

Mutual insurance: when output is different you can work together to guard against risk
(equal expected output but less risk).
If utility function is concave = risk aversion > people prefer mutual insurance.
More concave increases risk aversion
More risk lowers utility
The same average outcome but increase the spread; lowers utility.

Risk aversion
• Degree of risk aversion is determined by the curvature (‘concavity’) ofthe utility‐of‐money
curve u(y)
• Curvature is measured by the second derivative u’’(y)
• Arrow‐Pratt measures of risk aversion:
– Absolute: rA(y) = ‐ u’’(y)/u’(y)
– Relative: rR( )y = ‐ u’’(y )/u’(y ) * y

Mean preserving spread:

Suppose we have one lottery with outcomes x.
If we add another lottery z to this lottery where the expected outcome is 0 (on average you
add nothing).
Then the combined lottery x +z is more risky.

Find z such that x+z gives us the no insurance outcome.

Mutual insurance outcome: x > ¼=1000, ¼=1500, ¼ = 2000.

New lottery:
½ = -500 and ½ = 500
Average = 0 (-500+500=0 on average but increases the risk).

1/2U(1500)+1/2u(100)+1/2u(2000)

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Check that:
– Mutual insurance increases the expected utility
– There is no scope for mutual insurance if risks are perfectly positively correlated
– There is full insurance possible if risks are perfectly negatively correlated

Formal insurance
- Client is risk-averse (pays fixed premium to insurer)
- Adverse selection > asymmetric information (hidden information; need to make health
insurance obligatory).
- Moral hazard > conflict between insurance and incentives partial insurance as a
compromise
• hidden action problem; change behaviour after insurance is taken up (parial insurance =
compromise, some individual risk to ensure proper behaviour).
- Enforcement: usually through pre-payment of the premium

Mutual Insurance
1. Decreasing spread of potential outcomes increases utility if agents are risk averse
2. Insurance concerns unilaterial transfers that are unrelated to past or future transfers(<> not
credit!)
3. Power of mutual insurance depends on degree of correlation between outputs of agents
4. Mutual insurance is more viable if the number of agents increases and their fortunes are
less correlated
5. Informal mutual insurance usually based on social norms of reciprocity rather than formal
well‐specified contracts.

Works when agents are risk averse, if their degree of correlation is limited, if there are
enough agents involved in mutual insurance, usually use of social norms and reciprocity
(where formal insurance fails).

Two types of risk

- Idiosyncratic risk; individual level
- aggregate/ systemic/co-variant risk; societal level
Main point; lots of risks that can be insured; risks within the village.

Perfect insurance model:

Yt= A+e+o

Y=income of farmer i at date t

A= Farmer i’s average income
e = random shock to farmer i’s income at date t (idiosyncratic)
o = common shock to all farmers in the village at t (co-variant)

mean idiosyncratic shock is zero: Yt=A+o.

Y = insured income of farmer i at date t.

Testing the perfect insurancemodel

• Perfect insurance: hh level variables( income, shocks) should not affect hh consumption
• Estimate household consumption in deviation from village mean consumption: zie slide.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Conclusion;
consumption smoothing takes place; could be through mutual insurance but could also be
through self-insurance or credit market (but it is taking place!).
• Other studies: studies: general conclusion is that “full insurance” or “perfect risk sharing”
has to be rejected
Could be due to smoothing ex-ante; work on making their income stream more stable (plant
crops that are safer but lower return).

Why is insurance not perfect?

Problem with information (moral hazard) and enforcement (moral hazard).
- What is the final outcome? (how do you check how much the output from the harvest is?)
Can be improved through; social capital, information technology
- Actions that might lead to lower outcomes; farmers put in lower effort.

1. Problem of information, moral hazard

• How to give incentive for effort before harvest(ex‐ante)
A ‐ The final outcome : harvest, health/illness, etc.
• How to improve this information:
– Role of social capital
– Information technology
B ‐ What led to the final outcome
– Unobservable actions that lead to the actual outcome(moral h d)

Model: Limited information and Moral hazard

- L (Low) and H (High) output
- (uncertainty is idiosyncratic)
- Two probabilities of producing high output: p and q , with p > q, dependent on efforts and
input at extra cost C
- Probability of success depends on effort and inputs; p and q, and costs C.
With effort: pu(H) + (1-p)u(L)-C
Without: = qu(H) + (1-q)u(L)

• Assume that, without insurance, farmers will put in extra effort:

Pu(H) = (1-p)u(L) – C > qu (H) + (1-q)u(L)
• If farmers are risk averse: u(p) + (1-p)L) > pu (H) + (1-q)u(L)
• Perfect insurance:
- high output: pay (1-p)(H-L)
- low output: receive p (H-L)
• Each farmer receives the constant amount: pH+(1-p)L.

Viability of insurance if sufficiently large number of farmers: mean inflow = mean outflow
Viability constraint; production should be equal to consumption.

Incentive problem: payoff with low effort > payoff with high effort
Incentive constraint; make sure the farmers are willing to put in the extra effort.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Example p. 664
P=3/4, q=1/4, H=2000, L=1000, u=√ , c = 2.
Solve for x and y.
X = 1832, y = 1500.
Original spread between 1000 and 2000, perfect insurance is 1505, but with the incentive
constraint you can get to 1832.

2. Limits to insurance: enforcement

Lets say insurance is perfect but no way to enforce insurance.
One-time gain G from deviation from insurance scheme for farmer who produces high input
H (mean output Is M):
G=u(H)-u(M)

Loss L of deviation for each of the n periods after: L=N{u(M)-[pu(H)+(1-p)u(L)]}+S

Enforcement constraint for perfect insurance: L>G: N{u(M)-[pu(H)+(1-p)u(L)]} +S > u(H)-

u(M)

If this doesn’t hold; 2nd best:

Another nasty expression.

What can we say:

Left hand site (see slide 28); depends on X and Y but Y also depends on X (so really only
depends on x).
Right hand side: constant function (just a number = straight line in curve (depending on size
of S)).

Slope is positive; if x gets bigger, y has to become smaller (tendency to become more
negative).

Three cases:
- A = high social sanction; incentive constraint/enforcement constraint is always satisfied
- B = partial insurance; two possibilities, choose the lower one
- C = no insurance; people will always run away.

Intuitively
If the past payment increases pay-out in the future then the enforcement constraint becomes
less binding.

More informal insurance; harder too meet the enforcement constraint (loss from deviation is
lowered)
Aggregate shock; everyone is worse off, all incomes go down (gain from deviation is higher).

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

HC 10. Foreign Aid

Official development assistance (ODA) or Foreign Aid
- Can be bilateral or multilateral,
- There are two definitions:
1. Effective Development Assistance: grants from donor nation or institution and
2. Official D A: grants and concessional loans net of repayment of previous aid loans.

→ Data:
Private capital inflows are rising, increase in grants over years (decrease in loans).
Large regional variation: Africa still very dependent on aid, east asia not so much.

→ Official development assistance: norm: 0.7%, in reality around 0.25% is given (on
average).

Categories of ODA:
1. social (health, education etc.),
2. economic (production and economic infrastructure),
3. other (consumption in emergency situations)
Economic started out the biggest but resulted in the smallest; more focus on social stuff now.

Why aid?
Traditional argument:
- Financial: project finance for countries with poor access to world capital markets.
- Technical/knowledge (lack of knowledge): raising returns through donor’s role in project
selection and design (transfer of knowledge and money).
But: Many developing countries now have access to world capital markets
Fungibility: what you see is not what you get. hard to trace the actual impact of aid.

Policy argument for aid: structural adjustment programs; use aid to change policies
(conditions put on aid being given) conditionality
But:
- Ex ante conditionality (voorwaarden) doesn’t work; not done, adverse effects, under
pressure it does not work!
- Donors keep on disbursing (uitbetaling) aid when conditions are not met

Recipient has no incentive to maintain reforms if aid is temporary (time inconsistency) and
donor reluctant to punish reversals.
• Four possible outcomes if donor attempts to “buy” policy reform:
– desired policy changes not implemented (Zambia under Kaunda)
– implemented, but also in the no‐aid counterfactual (Vietnam,Uganda, Mozambique)
– policy change is due to donor pressure, but reversed (African trade reforms; Kenyan maize
marketing)
– policy change effected by donor and sustained

Making it effective is probably feasible, but not desirable. It would undermine:

– the incentives to develop capabilities for policy analysis,
– ownership,
– accountability

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

Selectivity argument for aid

– ex post conditionality: donors do not try to change polices but they are selective in
awarding aid.
• Under selectivity aid can have two new roles:
– signalling: if aid tied to success then private agents can economise onmonitoring. Rationale
for a donor role: underinvestment of private agents in information about small ,recently
reformed economies(e g.. African African risk ratings) ratings). selective and select
countries who have good circumstances for aid > then investors can use these choices to go to
where there is a positive development
– restraint: easier to resist pressures for policy reversal if aid allocation rule perceived as
credible. lose aid if you do not continue along the good governance path.

Realistic argument for aid: bureaucracy

The aid industry benefits from aid so there is little incentive to improve but just a self-
sustaining group (cynical argument).

International debate (books)

Sachs – poverty traps; financial problem
Ha-Joon Chang: bad Samaritans; policy argument = need good governance (we got rich
through protectionism and now we are telling developing countries that they cannot)
James Calderesi; selectivity argument didn’t work very well (aid given to people who are not
effective and only care about himself)
Paul cullier: bottom billion; aid effectiveness and globalization
Easterly: white man’s burden: main argument; development aid is organized top-down (this
idea that we know what is good for developing countries is incorrect, need more local
involvement).
Dambisa Moyo; dead aid = no government aid, some private aid (more financial markets).

Aid effectiveness
Hard to find a relationship on the macro level but good evidence of positive impact on micro
level.

Micro-Macro paradox – why?

- Aid is not large at the marco level (only less than 2% of GDP of low income countries)
- Project outcome variables are intermediate; support for economic development but not
economic growth
- Macro evidence is questionable; bad methodology and data
- Same for micro data; could be bad.
- Burnside and Dollar paper
- Aid has a positive impact on growth in developing countries with good fiscal, monetary,
and trade policies but has little effect in the presence of poor policies

Breakdown regression:
- Openness
- Inflation rate
- Budget surplus
- Added variable plot: two residuals (aid policy and growth); get exactly the same
coefficient for b5 if you get those residuals.

Downloaded by Gisha Bekele (gishabekele1@gmail.com)

 lOMoARcPSD|4405180

2‐‐gapmodel”
• Macro theory supporting aid (formany years): “2‐gapmodel” (Chenery (Chenery and Strout,
1966)
– First “gap” is between required capital to attain certain growth rate and available domestic
domestic savings
– Second gap is between import requirements for a given level of production and foreign
exchange earnings
• Foreign aid can fill the gaps
• Basis is the Harrod‐‐Domar growth model 1946

Harrod-Domar
g = (I/Y)/ mu
I/ Y = A/Y + S/Y
• µ = ICOR (incremental capital‐output ratio): additional capital required per unit of additional
output
• ICOR is often interpreted as a measure of ‘quality of investment’(high ICOR low quality)
2 key assumptions of the model:
• ICOR fixed over the short to medium run (Leontief‐ style production function)
• Aid is used for investment

Problems Problems with 2‐Gap /HDmodel (1)

• ICOR:
– Typically notfixed
– Reflects not necessarily quality of investment
• Reasons:
– Labor and capital are often substituted for each other depending on the resource
endowments(and markets/prices)
– In classic Solowmodel an increase in investment increases the ICOR (decreasing returns to
capital)
– In endogenous growthmodels: growth depends on many inputs(technology, human capital,
governance, quality, institutions) and they all affect the ICOR

Problems with 2‐Gap /HDmodel (2)

• Key assumption:
– aid is used to fill investment gap (A/Y = I/Y – S/Y)(and not used for consumption)
• Easterly: only correct if
– Investment incentives are right
– Financial constraints indeed binding
– (But if they were: why not financial inflows?)
• Empirical tests of financing gapmodel fail:
– Boone ( ) 1996:Aid finances consumption rather than investment in a cross‐country sample
– Easterly (2001): coefficient of aid in regression on investment is less than one for almost all
countries

Downloaded by Gisha Bekele (gishabekele1@gmail.com)