Transcript Document

The Household Economy
Approach (HEA):
‘regular’ and ‘lite’
The agenda:
1. Background: exchange entitlements.
2. The criteria for a useful method.
3. The method:
 getting the information
 using the information
4. Other uses and ‘lite’.
Until the1970s, the ‘cause’ of famine was still a
mystery. For example it was observed that:
 Despite widespread and sometimes deepening
poverty, famine was rare. In the 20 years to 1990,
& not including siege, perhaps 10 -15 cases,
some very local.
 Famine mortality was low – typically 3-5%
above expected mortality.
 Although most famines were preceded by
production failure, production failure rarely led to
famine. Some famines were not preceded by
production failure.
The explanation was provided in two
parts:

The observation that, when possible, people
took steps to protect themselves against
production failure and other shocks – so called
’coping strategies’.
1.
Sen’s (1981) theory of ‘exchange entitlements’
which provided a theoretical framework capable
of explaining all famines.
NDVI, Africa, June 1984
Source: NASA
‘Coping strategies’. People:
 keep food stocks, cash savings, livestock and
other goods, which can be exchanged for food.
 have systems of mutual insurance – ‘gift,
reciprocity and obligation’
and may be able to survive by:
 expanding existing sources of income e.g wild foods,
paid work, reducing non-food / food consumption
‘Exchange entitlements’:
A household (or person)'s ability to acquire food
depends upon its ‘entitlement set‘
 in practical terms essentially the sum of its current
income, food stocks, and the exchange value of its
labour and other assets.
This is independent of the adequacy of aggregate
food supply.
‘Coping strategies’ are a way in which households
seek to maintain entitlement.
The corollary to entitlement theory is that
the prediction of the effect of a ‘shock’ on the
ability of an individual or household to acquire
food requires a knowledge of that person’s or
household’s entitlement.
A difficulty:
 the value of entitlements may change as a result
of a shock e.g. the collapse of livestock/ labour
prices and a rise in the price of food.
Sorgum and goat prices, El Fasher,
Sudan, 1990
Sorghum
ep
t
S
Ju
ly
ay
M
ar
Goat
M
Ja
n
1200
1000
800
600
400
200
0
2. Criteria. An operationally useful
method is one which :
 can discriminate between and within populations
(the poor, better off) at an operationally useful level
of disaggregation.
 is practical i.e. can be used on large areas of diverse
economy, in war zones, at reasonable cost, by ‘non
specialists’
 provides output in terms which are useable by
governments and donors
 works in the context of the (usually) low quality of
official data on crop production and other shocks.
 works i.e. produces accurate output
The logical solution was to see if it was possible
to develop a model of rural economy based on
current quantitative economic descriptions,
obtained using ‘rapid’ field methods, to simulate
the relationship between a shock and an
outcome. Subject to:
the model being comprehensible to the user - not
a ‘black box’
This approach had the potential advantages of allowing
 complex multivariate problems to be managed in
an agreed framework, while retaining scope for
disagreement about matters of fact. Different users
might obtain a different estimates from the same data.
 for uncertainty in prediction – an analysis yields
a hypothesis which predicts changes which
should be monitored and observed if the
hypothesis is correct e.g. if an analysis
suggested that people will sell livestock people
should be observed doing this.
different interventions to be tested e.g. the effect
on access of holding down cereal prices
can be tested
Data required to estimate
household entitlement
The minimum data set includes:
 the household’s sources of food and non-food
(usually cash) income in a reference year.
 household assets - food stocks, cash savings,
livestock & other tradable assets.
 A description of the relationship between the
household and the wider economic context:
• the market for food, labour and other goods
• the ‘commons’ - hunting, fishing, wild foods
• other households - gifts & other ‘non-market’ exchange
For SCF, the central interest was to see if a model
could be developed which was sufficiently
sensitive to reliably predict the effects of shocks
‘short of starvation’ i.e. when households
 reduce non-food expenditure (education, health,
fuel, clothes) to meet food costs.
 are impoverished, as assets are exchanged for
food.
Getting the information:
The techniques have much in common with ordinary
survey techniques. An HEA assessment typically has 5
stages:
Defining the population(s)
Choosing the locations for interviews
Defining wealth groups at each location
Collecting household budgets for each wealth
group
Analysis
Populations are defined as:
 Populations of households which share a
‘reasonably similar economy’: boundaries
often follow agro-ecological zones.‘Reasonably’ is
defined by the purpose of the assessment.
II
IV
I
VI
VII
III
I1
D odoma
Dodoma
II
VIII
V
II
Sampling:
 Is purposive, not random
 Up to 10-15 sites may be visited in each defined
population
At each location in each area a wealth distribution
is established with key informants. The definition
of wealth is that used locally
Household interviews are conducted with
a group drawn from a wealth group:
 Allows people to discuss a ‘typical’ household
from their wealth group. This avoids the need to
discuss the income and expenditure of a specific
family.
Groups typically include men and women. In
addition, children are sometimes interviewed.
Calendar of agricultural and other productive
activities, Lower Kitui District, Kenya 2000. (SC
(UK))
The reference or baseline year
Group interviews refer to to a defined reference year.
 This may be an actual year, or (particularly in areas
where income fluctuates markedly from year to year)
a spread of years which, taken together, make a
‘reference period’.
 In areas subject to continuous disruption e.g.
parts of South Sudan, the reference year is
usually the year preceding the interview.
Data quality.
In any interview, income and expenditure should
balance.
 Food income should be in reasonable balance
with household energy requirements.
 Transactions – chiefly paid work and gifts –
should balance between suppliers (mostly
‘the rich’) and the recipients (chiefly ‘the
poor’).
Analysis:
Uses a simple arithmetic model to relate a
‘problem’ - a description of an ‘economic
shock’ to an outcome. The likely effect of the
shock on the ability of households in each
wealth group to acquire food and non-food
goods.
The model can be used by hand, on a
spreadsheet, or for large data sets using
purpose designed software - ‘RiskMap’.
Ranges
Data from household interviews in the same
population and wealth group are resolved into
ranges. For example:
 Barley production in the ‘very poor’ wealth group’
in population X is
35-45% of annual household requirement.
3 steps in analysis:
1. Defining the problem or ‘shock’
2. Calculating the immediate effect of the
problem on the ability of each household
type to meet its requirement for food and
non food goods
3. Estimating the ability of each household
type to compensate for any deficit in step 2.
Step 1: Defining the problem (1)
This has been the greatest difficulty. In most
countries:
 Crop statistics are, at best, rough estimates.
 Information on other types of production e.g.
rangeland conditions, fishing, is rarely gathered
at all.
 There may be little information on levels of
country/ regional food stocks, imports, & donor
intentions to supply food. It may be difficult to
estimate the market supply of food.
Step 1: Defining the problem(2)
The problem is usually defined from local
estimates of:
 likely production deficits, where possible
based on prior examples of similar e.g.
climatic, situations.
 probable price changes, given a particular
level of production.
Several ‘scenarios’ may be developed to cover
a range of possible situations e.g. with and
without price support.
Step 2. Calculating the deficit
For example, if a household usually gets 50% of
its income from food crops, a 50% fall in food
crops will lead to a 25% fall in household income.
Step 2. Deficit calculation
Expenditure
tobacco / drink
education
health
2500
diesel
firewood
2000
soap
matches
1500
dried fish
salt
1000
oil
cabbage
500
beans
sweet potatos
0
maize
rent
% HH Kcal Req/ year
Sources of food
100
50
Af ter 50%
rise in maize
price
Purchased
maize
Other f ood
purchase
0
green beans
-50
avocado
Huambo
2000.
Household
of 6
Step 3. Estimating the ability of a household
to overcome the deficit.
This requires making allowance for:
 Household food stocks (by addition).
 The possible expansion of income from wild
foods & gifts (arithmetic).
 Additional income from the sale of livestock,
labour and other non-food production (arithmetic
or using a linear supply/ price model.
Output
The aim of analysis is to develop an argument about the
likely effect of a shock, where the argument is :
 transparent and open to dispute
 based on a ‘complete’ information set and areas of
uncertainty e.g. the quality of specific data items,
are revealed.
 open to test and if necessary revision as
events develop and if predicted events do not occur
e.g. asset sales, reduction in meal frequency,
falling nutritional status
 The aim of HEA is to allow the user to think
through and systematically manage a complex
problem.
Where is HEA used? Who uses it?
SCF with WFP/FAO
National programmes
SCF
Method adopted by
FEWSNET,
SADC,WFP
+ many area/ refugee
assessments
+ Afghanistan
Costs
The largest cost is in establishing the baseline economic
description.
Maintaining a system is relatively cheap. The actual cost
depends chiefly on who is doing the work, being
comparatively higher where fieldwork is done by
expatriates: much lower when by national staff.
RiskMap:
% HH Kcal requirement
Predicted food needs, Malawi, 2002.
Using no ‘coping’ steps
16,000 - 29,000 MT
Using all ‘coping’ steps
10,600 - 18,000 MT
Estimated
deficit for
Malawi 2002 =
580,000 metric
tonnes
Other uses of HEA
1. HEA provides a way of measuring ‘vulnerability’
to food insecurity, where vulnerability is defined as:
the effect of a defined event (e.g. a price rise) on the
ability of a household/ population to access
sufficient food.
2. HEA has proved to be a very effective way of
teaching non-specialist staff the elements of
rural economy.
3. HEA has been used to evaluate food aid interventions.
4. HEA can be used to show the effect of other
economic changes on livelihoods. Trials are currently
underway in Kenya to see if HEA can effectively show
the effect of changes :
 to textile import tariffs on populations partly
dependent on employment in the textile industry.
 in international coffee prices on small scale producers
HEA ‘Lite’
Most of the time and cost of HEA is in acquiring
the data.
 Data can be obtained entirely from ‘key
informants’ with large savings in cost and time.
On current evidence (chiefly a national database
in Malawi) data quality suffers only in the loss of
detail. Where HEA is used to estimate large
changes e.g. crop failure, large price changes,
this is unimportant.
 HEA also provides a very useful framework for
rapid investigation of food security in e.g. single
households.