Wealth, Poverty, and Life Quality Dynamics:
Objective and subjective measurement and
interpretation based on a case study approach from Dinki
Alula Pankhurst[1]
Introduction
This paper attempts to analyze and compare selected households' recent histories in Dinki in Amhara Region,[2] studied by the Wellbeing in Developing Countries Project.[3] My aim was to understand their fortunes and their views of their standard of living and quality of life, particularly over the past ten years. I seek to explain why some household move into or out of poverty or wealth, whereas the status of others does not change, and consider the impact of events, shocks and social relations. I also try to understand how objective and subjective measures of their status and changes may be inter-related. The respondents' views of their happiness and the researchers' assessment of their contentment are compared with the household's wealth with a view to understanding how and why these vary by wealth category and over time.
This paper is based on a case study approach. The 34
households considered were selected based on data from the Ethiopian Rural
Household Survey (ERHS) for 1994 and 2004 and the 2005 Resources and Needs
Survey (RANS). Protocols and Instruments
were designed[4]
including open-ended questions, timelines and graphs. Interviews were carried
out with the household heads or their spouse by a female and a male researcher
in mid 2005.
The paper consists of five parts. The first outlines the selection of households and the methods used. The second considers the selected households according to five wealth categories. The third compares the households across the categories in terms of objective and subjective wealth measures and dynamics, standard of living measures and dynamics, happiness and contentment measures and contentment dynamics. The fourth part considers the findings in terms of the reasons for which the households were selected. The fifth concludes.
1. Modules, protocols, methods and household
selection
The households were selected based on three modules carried out in two rounds of research by the WeD Ethiopia team in June and July 2005. The first module sought to explore household histories.
The module had ten parts 1) open ended questions on the household's history over the past ten years, 2) members' history, 3) event history and an exploration of 4 key events, 4) a standard of living timeline of the household; 5) extra-household relations and activities, 6) intra household relations. module included 6 sheets 1) a Household Roster, to assess membership and changes, 2) an Events Calendar to record good and bad events over the past ten years, 3) a Standard of Living Timeline (SLT) on which to record changes since the household was formed, 4) a Standard of Living Graph (GHS), on which to place the household on the scale of 7 categories from the richest to the poorest used in the RANS and to plot changes from the imperial and Derg periods and especially the last ten years, 5) a Contentment Timeline (CT) on which to place changes in the household's contentment due to events, and 6) a Contentment Graph (CG) on which to plot changes in the household's contentment over the past ten years on a scale of 1 to 10. The second protocol was for ex-members, with the attempt to trace household members who had left, understand why they left, what happened to them afterwards, the consequences for the household and their subsequent relations.
The second module was based on the first adding a focus assets seeking to understand the impact of having or lacking assets among the wealthiest and poorest. The researchers were to check the assets listed in the RANS for the households selected according to very high or very low asset rankings and consider changes.
The third module was also based on the first with the
addition of a focus on social shocks which seem to be less studied in the
shocks literature on
The first module, carried out in June 2005, was based on 20 households selected from five specific types of households: 1) those whose objective and subjective wealth status was known to have changed between 1994 and 2004, 2) those who were known to be at the wealth extremes in 1994, 3) those whose perceived RANS wealth category in 2005 differs markedly from their current perception of their status five years earlier, 4) those whose perceived wealth category in the RANS is the same as their perceived status five years earlier, and 5) Those whose perceived RANS wealth category seems at odds with their perception of their happiness.
I. Households
whose status is known to have changed significantly over 10 years.
Six households: these were selected from households that were found[5] to have moved up or down by significantly[6] between the 1994 and 2004 ERHS in terms of three measures: consumption,[7] livestock[8] and self perception. Three households were selected from each category for each of the four rural sites. This paper only deals with the cases from Dinki though as we shall 12 further cases can be analysed as they were selected for other reasons.
II. The
richest and poorest households in 1994
Six households: three selected from the richest and three from the poorest based on the ERHS data. In each case the first was found to be the richest/poorest in terms of livestock, followed by land; the second richest/poorest in terms of land followed by livestock, and the third richest/poorest in terms of self perception, followed by livestock, then land.
III. Households
that in 2005 perceive their wealth
status as having changed over five years
Four households: two rich, one stating in the 2005 RANS that its status in comparison with the rest of the community has improved markedly, and the other claiming that its status has deteriorated markedly, and two poor, one claiming that its comparative status has improved markedly, and the other claiming that its status has deteriorated markedly.
IV. Households
that in 2005 perceive their wealth status
as staying the same over five years
Two households: one which states in the RANS comparative ranking that its status was rich and that its status has not changed as compared with five years ago, and the other which states that its comparative status was poor and has not changed over the past five years.
V. Households
whose wealth perception seems at odds with their contentment status
Two households selected from the RANS, one that states that it is very rich and yet has the low Global Happiness Score (GHS) of 'not so happy' and one that states that it is very poor, yet has the high GHS of 'very happy'.
It is important to note that the researchers were only provided with the household numbers and were told about the selection procedure but were not told which households corresponded to which categories in order to avoid their prior knowledge influencing the interviews.
The second and third modules were carried out in July 2005.
The second module was to be carried out with 6 households. The selection
criteria were that the household were among the top ten and bottom ten
households in terms of asset ranking based on an asset index derived from the
RANS. [9] The
asset index considered all the assets mentioned in the RANS. Indexes were
produced for rural and urban and for each site.
A select asset index was produced based on those that seemed most
relevant in each site.[10] Of
the six households three were to be selected from the top ten, and three from
the bottom ten, with a preference for those nearer the top and bottom. The
assets of the selected households in the RANS were to be checked and changes
noted and discussed.
The third module included 10 households that mentioned
social shocks in the RANS. These were to be selected from the following seven
categories[11] among
those that reported these shocks: 1) death, 2) accident/injury of family member,
3) disputes in family, 4) divorce of self, 5) dispute (contracts, inputs,
sales) 6) socio-political shocks (discrimination on a social or ethnic basis, land
confiscation, forced resettlement, bans on migration, conflict), and 7) positive
shocks for which remittances were selected. Each of the two researchers was to
interview 5 households from those identified in the RANS. Given the important
of death and divorce and the variety of socio-political shocks, two households
from each of these three categories were to be selected bring the total to 10
households.
1.1. The households
selected in Dinki
Overall in Dinki 36 households were interviewed by the male and female researcher in the two rounds in June and July 2005. However, two of those interviewed in the second round had already been interviewed by the other research in the first round for the household histories module, so that one household that was interviews on the grounds of being rich and unhappy had also faced the shock of death of a spouse, and the second was rich claiming improvement but was also interviewed as asset rich.[12] However, this gave the opportunity of different perspectives of the researchers and in one case a difference between the perspectives of the male head and his wife.
Out of the 20 households that were selected for the first module 19 were interviewed (the household identified as the poorest in terms of land and livestock in 1994) was not located and was substituted.[13] Although only 6 of the households were selected on the basis of the ERHS 1994-2004 comparison (three improving and three deteriorating), when we consider all the 34 households interviewed in the two rounds, 18 households were found to be among those identified by John Hoddinott as moving into or out of poverty. Of these 12 had improved and 7 had deteriorated.[14]
In the second round 16 households were interviewed, 6 for assets and 10 for shocks. Of the six households that were interviewed for assets, 4 were from the rich top ten (ranked 2,3, 4 and 5) and only two from the poor (ranked 5, and 7). However, out of the 34 households interviewed 9 were found to be within the top and bottom ten. Five of these were in the top ten (ranked 2,3,4,5, and 6) from the top and four were in the bottom ten (ranked 2,4,5, and 7) from the bottom.
As for the shocks households the 10 selected included two case of deaths (one of the male head and the other of his female spouse), one case injury, one case of dispute, two cases of divorce, one case of dispute (over land), two case of socio-political shocks (one of discrimination and the other of resettlement) and one case of remittances (that turned out to be a case of death of household head).
2. Discussion of the case
study households by wealth category
The 34 households were divided into five categories: 1) very rich, 2) rich, 3) middle, 4) poor and 5) very poor, based of a combination of measures, including livestock holdings, land holdings, their asset index, their self assessment and evidence from the interviews. Most weight was given to the 'objective factors', notably livestock and land. The decision where to place a household has been my subjective decision, though the range of indicators means that we have quite a good sense of the household.
Out the 34 households six were categorized as very rich,
five as rich, 11 as middle, five as poor and seven as very poor. The discussion
will consider each of these categories in turn regarding demographics, wealth
measures and dynamics, standard of living dynamics, global happiness and
contentment, and contentment dynamics.
2.1. Very Rich
households: Analysis
Out of the 34 case study households the following six can be
categorized as very rich (002, 036, 046, 049, 082, 129).
Demographics: All of the very rich households were male headed except for one woman whose husband had died in 2004. The ages of the household heads ranged from 30 to 80 (the widow being 30, and the next youngest being 45), with an average of 70.5 (or 78.6 if we do not consider the widow). The family size ranges from 3 to 8 with an average of 6.8.
Wealth measures: objective: The six households were characterized as very rich on objective grounds based largely on the RANS measures of land, livestock and asset ranking. All the households had more than one hectare of land, and 2 had more than two hectares, and the average was 2.29. All except one had irrigated land, and that case had the highest land holdings among the cases. All had two oxen and additional cattle, five had donkeys, and four had camels. Two each were in the 3rd and 5th asset quintiles and one each in the 2nd and 4th quintiles.
Wealth measures: subjective: In terms of their self evaluation in the RANS on a seven point scale from the richest to the poorest, only one described his household as 'among the richest', whereas 4 out of the 6 claimed to be 'about average' and one even claimed to be 'among the poorest'. We can therefore see a tendency not to acknowledge that they are among the richest.
Wealth dynamics: objective and subjective: five of the six households were in the EHRS sample, and four of these (002, 036, 046, 082), were found to have improved markedly, on all three counts: livestock, consumption and perception. The fifth one (049) was found to have declined significantly on two counts: consumption and perception, but not livestock. This was the case that claimed to be 'among the poorest' in the RANS. However, it is the richest of the cases in terms of land holding, and retains a significant herd including two oxen, and is in the 4th asset quintile. The sixth case was outside the EHRS sample and was the case of the widow whose husband was involved in trading using camels and was in the 5th asset quintile. The household had been improving until a recent sharp decline with the death of the household head.
Of the four households that were found to have moved out of poverty according to the ERHS data only one said that his household was 'much better' than five years ago, two said their household was 'a little better' and one said it was 'the same'. Compared to a year before the RANS, only one said it was 'a little better', a second said it was 'the same' and one said it was 'much worse'. The responses suggest some reluctance to acknowledge significant improvement. The household that was found to have declined according to ERHS data also said it was 'much worse' than five years ago and 'a little worse' than a year ago. This case was headed by an elderly man of 80 who was almost becoming disabled and almost bed bound.
Standard of living dynamics: subjective: In their RANS responses, compared with their father, three considered themselves to be poorer, two richer and one much richer.
Standard of living
dynamics: subjective and researcher: Compared to the imperial period, three
were richer and two poorer, and compared to the Derg period three were richer
and three poorer. Over the past ten years three improved and three declined.
The main reasons for improvement in wealth were said to be access to be good
prices for irrigated crops (mainly onions and fruit) especially in the last
three years, and involvement in trade using camels. The main reason for decline
seems to be the ageing of the household head and loss of labour with children
leaving.
Global happiness: subjective: Four of the six very rich households had a GHS score of 2 or 'fairly happy'. Only one had a GHS of 1 or 'very happy' and the widow had a score of 3 or 'not too happy'. This would suggest that it certainly does not follow that being very rich leads to great happiness.
Contentment: subjective + researcher: The score on the contentment graph in 2005 ranged from 3 to 9 out of 10. Only two scores were high (7 and 9) three around average (6, 5 and 5) and the widow had a low score (3).
Contentment dynamics: subjective + researcher: If we consider changes over the past ten years, three of the households were given constant scores, two at an average score of 5 and one at a high score of 7, whereas the other three households varied. Two of these dropped; in one case from high to average (8 to 6) as a result of ageing and deafness of the head, and the other from high to low (9 to 3) in the case of the widow. One case had a drop from above average to below average (6 to 4) when the head had trouble with a labourer, and then a steep increase to 9 out of 10 after they got another labourer. Whereas the contentment graphs show a more varied picture, with declines related to ageing, death of the head and problems with a labourer, we do not see sharp rises in contentment except in the case when a new labourer was found. The graphs therefore do not suggest that the very rich have been becoming happier.
If we consider the four household that have 'moved out of poverty', three had GHS of '2' or fairly happy. Two of these had constant graphs, one at an average of 5 and the other of 8, one had declined from 8 to 6 due to old age, and one had moved down and then up over the question of labourers. However, three of the four were above average, and the one that had an average score of 5 had been based on an interview with a wife who was not very happy with her husband due to his drinking, whereas he was the only one who had given a score of 'very happy' for the GHS.
2.2. Rich households: Analysis
Five households can be categorized as rich (020, 028, 079, 159 and 127).
Demographics: The rich households had heads whose age ranged from 30 to 70 averaging 47.2. The household size ranged from 2 to 6 with an average of 4.6. The therefore tend to be headed by younger household heads than the very rich and have somewhat smaller household sizes.
Wealth measures: objective: All except one had irrigated land, and the exception had quite a large land holding of 1.5 hectare. Four of them had two oxen and the fifth had one ox; two had 2 cows and two had 1 cow, and four had donkeys, whereas none had camels. Two were in the 5th asset quintile, 2 in the 2nd asset quintile and 1 in the 3rd asset quintile.
Wealth measures: subjective: In the RANS four out of the five claimed they were 'about average' and one claimed he was 'a little poorer than most'. This case (020) was clearly among the richest in terms of land holdings with 2.25 hectares, although its livestock profile was not among the richest with 1 ox 1 bull, 2 cows and 3 heifers, and it was in the 3rd asset quintile. The evidence suggests a distinct reluctance by the rich to see themselves as among the rich.
Wealth dynamics: objective and subjective: three of the five households were in the ERHS sample. All three were found to have improved significantly. One (028) was found to have improved significantly on all three counts: livestock, consumption and perception, one (079) in terms of consumption and livestock but not perception and the third (020) in terms of perception and consumption but not livestock. In the RANS comparison with five years ago the first of these (028) was said to have become 'a little better'; the second (079) was said to be 'the same', which seems to fit with the head's lack of perception of improvement, and the third (079) states that is it 'much better' off; though it is noteworthy, given the lack of improvement in livestock in the ERHS comparison that it is the only rich household without two oxen (it had 1 ox but also a bull, 2 cows and 3 heifers). The ERHS and RANS perceptions seem on the whole to match in suggesting improvement, except for 079 where the RANS perception of no change fits with the lack of improvement recorded in the ERHS.
Compared with one year ago the RANS responses of the first and second of these cases was 'a little better', and the third 'much better'. For the other two cases we only have subjective measures: in one case (127) the household was said to be 'a little worse' than five years ago, and in the other (159) it was seen as 'the same'. The former claimed to be 'the same as last year' but the latter 'much better'. The former had suffered two death of a daughter in 2001 and the wife in 2003. The latter was a young household only formed in 2003.
Standard of living dynamics: subjective: In the RANS comparison with their father, two household heads saw themselves as poorer, two richer and one the same.
Standard of living dynamics: subjective + researcher: Compared to the imperial period, two households considered themselves to be richer and one poorer, and compared to the Derg three considered themselves to be richer, and one poorer. Most of the richer households that were already formed under previous regimes seem to have improved. Over the past ten years, two households improved, one declined and one stayed the same; among the households two started from 'about average', one moving to 'poorer than most' and the other to 'richer than most; one remained 'poorer than most' and one moved from poorer than most to 'about average'. Comparison was not available for one household. Reasons for decline included livestock loss due to drought and disease, labour shortage with children leaving, (028), decreasing agricultural output and increasing number of household members (079), production shocks with crop and livestock losses and shortage of labour with a daughter leaving and the wife dying (127). Reasons for improvement included investing money from weaving to finance sharecropping (020), good parental endowments and involvement in the Kebele (159).
Global happiness: subjective: Two of the households (028 and 159) had GHS of 1 or 'very happy'. The former is a middle aged male household head who was found to have improved on all three counts in the ERHS. The latter is a young household head recently married and well endowed by his parents; his wife is pregnant and he is involved in the Kebele administration. Two others (020 and 079) have a GHS of 2 or fairly happy'. One (127) had a GHS of 3 or 'not so happy'. He is an elderly man of 70 who lost a daughter in 2001 and his wife in 2003, and suffered crop losses in 1999 and cattle losses in 2002. The evidence does not suggest that being rich necessarily implies being very happy.
Contentment: subjective + researcher: The CG scores for 2005 range from 2 to 6 out of 10. Three of these are very low at 2 out of 10 and only one is above average (6/10). Data is lacking for one case. The evidence suggests that most of the rich households are in fact unhappy.
Contentment Dynamics:
subjective + researcher: All four rich households for which CGs were drawn
show changes, but only one (079)
improved from a low contentment of 3 to just above average at 6. The
earlier low score might be related to problems the head had with the
administration and the improvement due to good cooperation within the
household. The other three households decline from about average to low (5 or 6
to 2). Reasons include the loss of the wife and daughter compounded by
production shocks (127), the wife leaving (028) with a slight improvement after
she returns, conflict with a labourer and possibly also between the spouses (020).
The evidence suggests that social factors, notably relations within the
households and social shocks relating to death and departure of a spouse,
explain the declines. Two of these households were found to have improved in
objective terms. One can therefore conclude that improvement in wealth does not
necessarily to go together with improved contentment.
2.3. Middle wealth households: Analysis
The following 11 out of the 34 case study households may be categorized as middle wealth (001, 009, 027, 057, 058, 065, 101, 116, 124, 174, and 176).
Demographics: The ten households categorized as rich include four female headed households. The age of the heads ranges from 29 to 70 with an average of 55.6. The household size ranges from 3 to 5 with an average of 4.4.
Wealth measures:
objective: In terms of land holdings, eight out of 11 had more than 1
hectare of land, four had more than 1.5 hectare, and one had more than two
hectares. What may be more significant is that the majority of the middle
income households (7 out of 11) did not have irrigated land. In terms of
livestock only one had two oxen, but six
had an ox, and two had half an ox; only three had a donkey. In terms of asset
quintiles, 4 each were in the 3rd quintile and 5th
quintiles, and one each in the 4th 2nd and 1st
quintiles.
Wealth measures:
subjective In terms of their RANS comparative self assessment, four of the
households considered themselves to be 'about average', three each 'a little
poorer than most', two 'among the poorest', one as 'the poorest' (a female
headed household) and one as 'richer than most' (this case 176 had a hectare of
land and 1 ox, but was in the 5th asset quintile). Half of the
middle households therefore consider themselves to be poorer than average, and
only one sees himself as above average.
Wealth dynamics: objective and subjective: Four of the ten households were among the EHRS sample that were found to have changed status. Two of these (001 and 101) were found to have become significantly poorer in terms of perception and consumption but not livestock; the other two had become significantly richer, in one case (057) on all three counts, and in the other case (027) in terms of consumption and livestock but not perception. Regarding their RANS responses to the comparison with five years ago, one of those that was found to have become poorer in the ERHS also said it was 'a little worse' than five years ago (001) whereas the other said that it was 'the same'. As for the two who were found to have improved, one (057) said it was 'a little better' whereas the other said it was 'the same'. However, compared with a year earlier the first of the four (001) said it was 'the same'; the second (101) said it was 'much better; the third (057) 'a little better' and the fourth (027) 'the same'. The responses of the fourth case (027) in the RANS questions were both 'the same' and may explain the fact that the head did not mention an improvement in the ERHS. This household head had been in power under the Derg and lost influence since; he also faced problems with his children, with a daughter converting upon marriage and then divorcing and a son dying, poor health with ageing and having to look after grandchildren. It seems that his personal and family situation may have affected his evaluation of the improvement in wellbeing.
Global happiness: subjective Most of the middle households (7 out of 11) had a GHS of 2 or fairly happy. Two households had scores of 1 or 'very happy' and two others of 3 or 'not so happy'. One of the households that gave a 'very happy' score (124) was a relatively young male head who was happy with the second wife who works hard, and inherited land from his father. The other (009) was an older man who also had a divorce and was happy with his new wife, who is treating his children well. One of the 'not so happy; households (057) is a female headed household whose daughters have been marrying out with one returning divorced with a child. The other (027) was unhappy with a daughter changing religion and a son dying, and having to look after grand-children as well as with his loss of status. The evidence regarding the extremes in happiness suggests that spousal relations are key in great happiness and poor relations with children in accounting for discontent.
Contentment: subjective + researcher: CGs were produced for eight households; they ranged from 1-8. Three of these were average at 5 out of 10; three were high at 7 and 8, one was just above average at 6, and only one was below average and low at 1 out 10. This case is an elderly female headed household who has become poorer, who lost a daughter in 2000 and who relies on a son who she does not consider is a hard worker, and has to look after an ageing mother. The majority of the middle level households have average or above average GH and CG scores
Contentment Dynamics: subjective + researcher: Of the nine households with CG scores, five had remained constant (three of these at an average level of 5, one at a high level of 7 and one at the very low level of 1 out of 10 (the female headed household mentioned above). Of the remaining four one (176) had simply risen from 7 to 8 out of 10 when the young head's first child was born. Two others had started below average fallen further and risen above average. Both these cases were associated with divorces and remarriages. In one case (009) the fall to rock bottom was associated with a divorce of an elderly man and a steep rise to 8 out 10 with his remarriage and happiness with the new wife who was treating his children well. In the other case (065) the drop was to 2 out of 10 at the time of his second divorce, but the rise was only to just above average, no doubt because the head was not fully satisfied with his third wife and regretted divorcing his first wife. The last case (116) had declined sharply from 8 to 2 when his wife died with a further decline to 1 which may have to do with unhappiness with his new wife. The evidence suggests that decreases in contentment are associated largely with divorces, and increases with remarriages and in one case the birth of a first child. A less significant recovery or further decline after death or divorce of the wife was associated with dissatisfaction with the new wife. This suggests that family and especially spousal relations are the key factors in contentment dynamics among middle households.
2.4. Poor households: Analysis
The following five out of the 34 case study five households may be categorized as poor (023, 025, 056, 086, and 125).
Demographics: The five households include one female headed household. The age of the heads ranges from 19 to 80 with an average of 45.4. The household size ranges from 1 to 5 with an average of 3.2.
Wealth measures:
objective: In terms of land holdings, none had more than two hectares, two
had more than 1.5, and only one (the female headed household) had less than one
hectare. One had irrigated land. In terms of livestock none had oxen. In terms
of asset quintiles, two each were in the 3rd and 1st
asset quintiles, and one was in the 2nd quintile.
Wealth measures: subjective In terms of their RANS comparative self assessment, three considered themselves 'among the poorest' and two 'a little poorer than most'. We therefore see fairly good fit between the objective and subjective measures.
Wealth dynamics:
objective and subjective: Two of the five households were among the EHRS
sample that were found to have changed status. One of these (025) was found to
have moved into poverty on all three counts, whereas the other (086) have moved
out of poverty on two counts: consumption and livestock, but not perception. Regarding
their RANS responses to the comparison with five years ago, the former
suggested that it was 'a little worse' off, whereas the latter suggested it was
'the same'. These responses suggest a good fit between the ERHS and RANS
perceptions, as the former case suggests getting worse and the latter does not
acknowledge improvement.
Standard of living dynamics: subjective: In their RANS responses, compared with their father, all the household considered themselves poorer and one of them 'a lot poorer'.
Standard of living
dynamics: subjective + researcher: Compared to the imperial period, the
only case which was relevant was richer, and compared with the Derg three of
those that had households at the time were richer. Over the past ten years
three of the household declined and only one (023) improved very slightly over
the past couple of years within the 'a little poorer than most category' owing
to good harvest, involvement in daily labour, the cow his new wife brought and
her income from cotton spinning . The main reason for decline seems to be the
ageing and disabilities of the household head and death of the household head
followed by household disintegration, and loss of labour with adult children
departing.
Global happiness: subjective:
Three out of the five households have GHS of 2 or 'fairly happy'. One (056)
has a score of 1 or 'very happy' and the last one (086) a score of 3 or 'not
too happy'. In the former happy case the head is a young man whose mother divorced
his father and brought the children to Dinki to live with her brother who died
leaving them land. The mother also died, some sisters went to live with his
father, and others remained with him. He recently got married and this might
explain the high GHS. The unhappy case is an elderly man of 70 who had become
blind in 2004 stays and home and was feeling unable to help his household.
The evidence regarding the extremes in happiness suggests that marriage of a young man can result in happiness (despite the households demise with the death of the mother), and that ageing and disability account for discontent.
Contentment: subjective + researcher: CGs were produced for all the five households and ranged from 1 to 9. Only one of these (125) was average at 5 and remained constant. All the others changed as we shall see. Two ended up well above average at 7 (086) and 9 (023) out of 10, whereas the other two ended up very low at 3 (025) and 1 (056) out of 10. The evidence suggests a wide variation in contentment among the middle wealth category.
Contentment Dynamics: subjective + researcher: Only one of these was average at 5 out of 10 and remained constant. This is surprising as the household (125) is composed of a single woman who we know to have been unhappy. She was married against her will by her father, and her husband took her to resettlement against her will. They returned and she divorced him while pregnant and went back to her father. She quarreled with her step mother and set up her own household. One might interpret the average constant score as reflecting the difficulty in assessing her status as she was not a separate household (her SLT starts in the last two years), an unwillingness to give a negative score or an average score in fact being negative. Of the remaining cases three had decreasing contentment. One of these (086) had a high contentment of 8 out of 10 which only dropped slightly to 7 out of 10 in the last year. This case is an elderly man of 70 who lost his sight in 2004 and became homebound. The other two showed greater declines. One of these moved from average (5 out of 10) to very low (1 out of 10). This case (056) is a young male head whose mother divorced his father and brought the family to live with her brother in Dinki. Both the uncle and mother died and some of the siblings went to live with their father. The second case (025) declined from a high level of 7 out of 10, to below average 4 and then further to 3 out of 10 in the last two years. The head is an elderly man of 80 who lost his hearing and more recently his sight and is facing disabilities. The final case moved down and then up (from 4 to 2 then 8 and 9). This case (023) is a young man who got divorced and then lived alone at the time of low contentment, and remarried and had a daughter which coincides with the increasing contentment. The evidence suggests that ageing and disability, notably blindness and death of a parent and household disintegration are reasons for decrease in contentment. A decrease followed by a sharp rise was explained by divorce followed by remarriage.
2.5 Very Poor households: Analysis
Out of the 34 case study households the following seven
can be categorized as very poor. (042, 066,
069, 071, 074, 184, and 190).
Demographics: most of the very poor (four out of seven) were female headed. The ages of the household heads ranged from 20 to 80 with an average of 53.1. Both the young cases in their twenties were men, and the average age for the female heads was 67.7 years. The household size ranges from 1 to 3, with an average of 1.8 members. Four of the households were single heads and the other three had 3 members each.
Wealth measures: objective: None of the households had more than a hectare of land, three of them had less than half a hectare and two had none. Only two households had access to irrigated land. None of the households had oxen. In terms of assets six were in the first quintile and the last one curiously was in the fifth asset quintile.
Wealth measures: subjective: In terms of their subjective self evaluation in the RANS on a seven point scale from the richest to the poorest, four described themselves as 'the poorest', two as 'among the poorest', and one as 'a little poorer than most'. We can therefore see that the very poor tend to see themselves as such.
Wealth dynamics: objective and subjective: Four of the seven households were in the EHRS sample. Three of these were found to have declined markedly. Of these one (071) declined on all three measures, whereas the other two (042 and 069) declined in terms of perception and consumption, but not livestock. The fourth case (084) was found to have improved in terms of perception and consumption, but not livestock. This would suggest that livestock are difficult to measure for the very poor, presumably since their holdings are too small. Indeed if we look at the livestock holdings of the three cases that declined they are meagre: 071 had no livestock, 042 has only 3 goats and a chicken, 069 has only six chickens. As for the fourth case, she had a shared cow and a goat.
Of the three households that were found to have moved into poverty according to the ERHS data the two female headed households said they were 'much worse now' than five years ago, though the male headed household (042) curiously said that he was much better than five years ago, although he placed his household as 'the poorest' in the community. This may be because since the death of his father and the departure of his mother with two of his brothers he has become the household head. This might suggest that despite the decline in objective and in subjective terms stated in the ERHS, the household head one year later feels that his independent status makes him better off. It may also be that since his mother and siblings left (one sister who married away came back), he has not got responsibilities and less mouths to feed. The female household head whose household had improved in terms of consumption and perception in the ERHS suggested that her household was 'the same' in standard as five years previously. However, the SLG does suggest an improvement in the EPRDF period. The head has engaged in daily labour and has land she gives to sharecroppers and benefited from the sale of irrigated onions, and was able to obtain a share of a cow.
Standard of living dynamics: subjective In their RANS responses, compared with their father, all the households considered themselves to be poorer and two a lot poorer.
Standard of living dynamics: subjective + researcher: Compared to the imperial period, in the SLG four of the elderly household heads said they were richer in both imperial and Derg periods. One female headed household (084) came from a poor household and was marred off at 15. One young man (190) came from a poor household and poor parents who were land-owners during the Derg; due to their poverty he became a servant in rich households. The major reason for decline seems to be ageing, disability and limited support with children leaving or unhelpful (069, 071, 074). Other reasons include loss of assets in a fire (071), theft of livestock (074), land returned due to disability and the remaining plot given to sharecroppers and subject to tax (066), and for female headed households death of the household head (071), giving out land to sharecroppers (069) and land confiscation (071, 074).
Global happiness: subjective: Four of the seven very poor households had a GHS score of 3 or 'not too happy', and the remaining three had a score of 2 'fairly happy'. The latter included a young man (190) who had been a servant and recently established his own household, got married and had a daughter, a middle aged woman (084) whose household was found to have improved in the EHRS, and whose CG rating is very high at 7 out of 10. She improved her life from wage labour, gained access to irrigated land and a share in a cow. The third case (071) is more difficult to interpret. She is an elderly female single woman of 72, whose husband and only son died. Her son's children do not help her. She expressed unhappiness with ageing and also has a sharply declining CG. Perhaps this elderly woman did not feel comfortable expressing dissatisfaction. She is a Muslim and researchers suggested that the Muslim Argobba are less willing to express openly dissatisfaction, as there is a feeling that Allah expects people to praise him and mention contentment.
The evidence suggests that the very poor are certainly not very happy. However, three of them suggested they were fairly happy, two of these having had improvements in their circumstances, one having established his own household after being a servant and the other gradually improving her life through income from wage labour. The third case seems to be reluctance by an elderly religious Muslim woman to express unhappiness openly.
Contentment: subjective + researcher: The score on the contentment graph in 2005 ranged from 1 to 10; however apart from two cases, they were all well below average. The two exceptions are the case of the female headed household mentioned above whose livelihood is improving (084) who had a constant score of 7 out of 10, and the former servant (190) who had a score of 10 out 10 but only for the last two years since he set up his own household with the graph not filled in for the earlier period while he was a servant. The remaining cases had scores of (4,3,1 and 1).
Contentment dynamics: subjective + researcher: If we consider changes over the past ten years, apart from the two cases of high contentment mentioned above (084 and 190), all the other cases were declining. from about average to low: The biggest drop was from 6 to 1 for the elderly single female headed household (071) who lost her only son and had no support, followed by a decline from 5 to 1 of another elderly female headed household (069), who lost her hearing and sight, followed by a decline from 6 to 4 of a third elderly single female headed household (074) whose granddaughter left and suffered from loneliness. The only single elderly male head (066) also suffered a decline from 4 to 3. He had no support and gave out his small plot to sharecroppers and was upset at having to pay taxes on it. The evidence suggests that most very poor, especially the elderly and women headed households have contentment graphs that show a marked decline from about average to very low. This is due to ageing, disability, notably blindness, lack of support, departure of kin, and in the case of the elderly single man inability to cultivate the little land they have.
3. Discussion comparing the wealth categories
The following section draws conclusions from the
preceding section comparing the evidence for the different wealth categories in
terms of demographics, wealth measures, wealth dynamics, standard of living,
global happiness, contentment and contentment dynamics.
3.1. Demographics: household size, age and gender
of household head
Table 1: Wealth
categories and household size, age and gender of head
|
|
number of
households |
number of female
heads |
age range of head |
average age of head |
household size
range |
average household
size |
|
very rich |
6 |
1 |
30-80 |
70.5 |
8-10 |
6.8 |
|
rich |
5 |
0 |
30-70 |
47.2 |
2-6 |
4.6 |
|
middle |
11 |
4 |
29-70 |
55.6 |
3-5 |
4.4 |
|
poor |
5 |
1 |
19-80 |
45.4 |
1-5 |
3.2 |
|
very poor |
7 |
4 |
20-80 |
67.7 |
1-3 |
1.8 |
|
total / average |
34 |
10 |
19-80 |
57.3 |
1-10 |
4.2 |
Table 1 shows that the highest proportion of female heads is in the very poor category, where they are more than half. The only female head in the very rich category is a recent widow. There are, however, 4 middle wealth women headed households. The age range of head is from 19 to 80 and the average age is 57 years. The higher ages are among the very rich and the very poor. The household sizes range from 1 to 10, with and average of 4.2 persons. There is a clear tendency for household size to decrease from the very rich with an average of 6.8 members to the very poor with an average of 1.8 members.
3.2 Wealth measures: objective
The households were ranked mainly in terms of the land and livestock holdings. There was therefore a rough correlation in this respect.
land holdings: The very rich households had more than 1 hectare of land, with an average of 2.29, and all except one had irrigated land. The rich households apart from one had over two hectares with an average of 2.18, and only one did not have irrigation. The medium households had an average of 1.42 hectare, but 7 out of the 11 did not have irrigated land. The poor households had an average of 1.27 hectare, and only one had irrigated land. The very poor households had an average of 0.34 hectare and only two out of 7 had irrigated land. As would be expected there is a clear increase in land holding with wealth and a greater proportion of households with irrigated land.
Livestock holdings: Using oxen as a proxy, the very rich have two oxen, the rich apart from 1 have two oxen (an average of 1.8), only one of the middle has two oxen, and five have one, the average being 0.64; and none of the poor or very poor have oxen. Five of the very rich had donkeys, one a mule and 4 had camels, and none of the others categories had camels or mules. The very rich had 1-3 cows except for one case, the rich had 1-2 cows except for one case, most of the middle had one cow, only three of the poor had a cow, though two others shared a cow, and none of the very poor had a cow.
Asset quintiles: The asset quintiles and ranking did not match very closely with wealth as considered in terms of land and livestock. However, none of the very rich or rich were in the lowest quintile, whereas all the very poor except one were in the top quintile, and two of the poor were in the top quintile and the rest were in the second and third quintiles.
3.3 Wealth measures: subjective
RANS self placement: On the scale of 1 to 7 from richest to poorest, there was a tendency of the richest and rich not to describe themselves as such. Only one of the richest described himself as 2 or 'among the richest', and four as 'average' and one even as 'among the poorest'. The average score for the richest was 'about average'. Likewise four out of the five rich described themselves as 'average' and one as 'a little poorer than most'; the average score for the rich was also 'about average'.
Among the middle wealth category, out of 11 households four considered themselves 'about average' three 'a little poorer than most', two 'among the poorest' and one as 'the poorest'. The average was close to the fifth category 'a little poorer than most'.
In contrast the poor and very poor were keen to emphasize their poor status: three of the poor considered themselves 'among the poorest' and only two 'a little poorer than most'. Among the seven very poor four described themselves as 'the poorest', two as 'among the poorest' and one as 'a little poorer than most.
3.4 Wealth dynamics: objective and subjective
ERHS/RANS comparisons
In the total ERHS
sample 50 households were found to have changed status significantly. [15]
Of these 23 were found to have moved into
of poverty (6 on all three measures, 13 in terms of perception and consumption,
4 in terms of consumption and livestock and none in terms of perception and
livestock). 28 households were found to have moved out of poverty 11 on all three measures, 4 in terms of perception
and consumption, 5 in terms of perception and livestock, and 8 in terms of
consumption and livestock.
Out of the case study households in the WeD sample 18 were found in the EHRS comparison between 1994 and 2004 to have changed status significantly. Of these 7 had become significantly poorer, and 12 had become significantly richer.
3.4.1. 'Moving
into poverty' or getting poorer
If we consider first the six[16] that had moved into poverty, three of these (049, 071, and 069) also said that they were 'much worse' than five years ago, in other words the bottom score of 5 out of 5. One (025) considered that they were 'a little worse', one (101) considered that she was the same, and another (042), thought that he was 'much better now'. There seems therefore to be a fairly good fit between the EHRS data and the RANS perceptions. Let us consider the two exceptions:
The first (101) is an elderly
household head with an ageing mother, a son aged 20 and another aged 10. In the
RANS she had two hectares of land, an ox, a cow, a heifer and a calf, which may
explain why she is seen as moving into poverty in terms of consumption and
perception but not livestock. She also suggested that she was 'much better' off
than a year earlier, which may suggest that there has been some recent
improvement which had not fully taken effect at the time of the ERHS survey but
which influenced or reduced the sense of decline. However, she does rank her
household as 'the poorest' and poorer than her father.
The second case (042) may be
somewhat more straightforward. The head is a young man whose father died eight
years ago and whose mother left with his siblings. One sister married but
returned divorced. Both of them work as labourers. Although he too places his
household as 'the poorest' it may be that having his own independent household
and beginning to get over his bereavement which would suggest some improvement.
However, it is noteworthy that he claims no change since a year earlier. In the
ERHS data he was found to have become poorer in consumption and perception but
not livestock though his current livestock (only 3 goats and a chicken) make
the household very poor and he obtains oxen through exchanging labour.
Looking at longer term trends, in terms of the subjective
comparison with their fathers five out of the seven for whom data is available
see themselves as poorer than their father. It is noteworthy that two of the
households moved into poverty on all three measures, whereas four did so in
terms of consumption and perception but not livestock. Of these two had no
oxen, one had one ox, and one had two oxen.
It is also important to note that although we have been talking of moving into poverty, in fact out of the 7 cases, only three (042, 069 and 071) have been categorized among the very poor in this study and consider themselves to be either 'the poorest' (071 and 042) or 'among the poorest' (069), and one (025) was categorized as poor and considered himself as 'a little poorer than most'.
Another (101) was considered a
middle wealth households with one ox, a cow and a calf and more than a hectare
of land, although she described herself as 'the poorest'. This households also has a relatively good
asset standing and was in the 4th asset quintile ranked 24th.
Since this household was found to have declined in perception and consumption
but not livestock, this might suggest that the negative perception is affecting
a view of decline. It is also noteworthy that the household has grown up sons.
The remaining households (049)
is in fact in the rich category with 3.25 hectares of land and 2 oxen, a cow, a
calf, a donkey and 14 goats. He was in the 4th asset quintile ranked 23rd.
However, he ranked himself as 'among the
poorest', and ' much worse off' than five years ago and 'a little worse off
than a year ago. The head is 80 years old, getting weak and bedridden and hardly
able to talk and his wife who is 85 is deaf and unable to go out. The decline
from very rich even if significant does not mean moving into poverty, as the
household still retains enough land and livestock, support from a labourer and
a son with his own household. However, the ageing of the head may also have
resulted in a reevaluation as much poorer than the objective data suggest.
The last two cases discussed suggest that only some of the
households that have declined have actually moved into poverty, whereas others
have become poorer. A high weighting for the perception that might be
influenced by other factors such as ageing and poor social relations, could
make households that are relatively well off appear to be poorer than they are.
3.4.2. 'Moving
out of poverty' or getting richer
When we consider the 12 cases that have moved out of poverty, two (046 and 020) say they their household is 'much better' off than five years ago, three (028, 082 and 036) say they are 'a little better', but five (002, 027, 079, 084, and 086) say they are 'the same', and two say they are worse off (001 and 057).
The last case (057) is a
middle-aged female head who considers herself to be 'a little poorer than
most', but was categorised in the middle wealth category. However, she views
herself as 'a little better' than a year
ago. The household had a sizeable amount of land at 2.63 hectares of land;
however none of it was irrigated. In terms of livestock they had 1 ox, 1 bull,
1 cow, 1 heifer, 1 calf, 2 donkeys, 6 goats, and 14 chickens. It was in the 5th
asset quintile, ranked 26 out of 34. She
views herself as poorer than her father. As we shall see the household head is
unhappy with her social relations, notably her daughters leaving and this may
have influenced a rather negative self perception.
In terms of longer term trends in perceptions comparing themselves with their father seven of the households consider that they are poorer than their father, and only four that they are richer. Four of those who considered themselves poorer had improved in terms of consumption and livestock but not perception, and two of these (027 and 086) viewed themselves as 'a little poorer than most' one (001) as 'among the poorest' and the third (079) as about average. This might suggest that a negative self perception is affecting the view of their wealth status.
Considering the wealth ranking of those who have moved out of poverty, out of the 12 four (002, 036, 082, and 046) were among the very rich, three (028, 079 and 020) were among the rich, three (057, 001, and 027) were among the middle, and one each were among the poor (086) and the very poor (084). Let us consider the last two cases.
The former (086) is a
household headed by an elderly man who is becoming disabled and blind and now
stays at home. The household seems to be relatively poor. It had enough land at
1.56 hectares, but not much livestock (a shared cow, a shared bull and a shared
calf, and 1 donkey). It was in the 1st asset quintile ranked 7 out
of 34. The head ranked the household as 'poorer than most' which may be
accurate. The EHRS shows an improvement in livestock and consumption but not in
self perception. The RANS self perception fits with this in that it suggests no
change in the past five, although it does mention a marked improvement in the
past year. The head's ageing and loss of sight of the head may explain the
differences between the objective and subjective data. The fact that the
household is at the end of its cycle with a son who has established his own
household and the head ageing may explain the perception of decline despite apparent
evidence of improvement.
The latter (084) is a female
headed household who despite being poor had used income from wage labour as a
forest guard and a cook and a small irrigated plot planting onions to improve
her life. She has shares in a cow and has a goat and has good relations with
her neighbours. Although she is clearly very poor, and was second in the asset
ranking, she only placed herself as 'a little poorer than most', rather than
'among the poorest' or 'the poorest'. This might suggest that her feeling of
independence and small improvement have given her a sense of relative success.
She was found to have improved in both perception and consumption but not livestock.
The discussion of the last two cases may explain why
households that have been found to improve in the ERHS seem to be poor. In one
case the head's age and disability may have coloured his perception and gave a
sense of decline. Indeed, his household was found to have moved out of poverty
in terms of consumption and livestock but not perception. The other case is a
fairly young female headed household who was seen as moving out in terms of her
perception and consumption but not livestock. Although she is clearly rather poor,
her sense of independence, small improvement and perception of herself as only
a little poorer than average may have led to a higher evaluation of
improvement.
3.5. Standard of living dynamics
According to the RANS question regarding how they compare themselves with their father at the same age, the majority, 20, consider themselves poorer and a further three consider themselves 'a lot poorer'. Two consider themselves 'about the same' and only six consider themselves 'richer' and none consider themselves as 'a lot richer'. Of the three that consider themselves as 'a lot poorer' two are in the very poor category and one in the poor. Of the six that consider themselves 'richer' three are in the very rich category, two in the rich and one in the middle. Three of these were found to have improved in the ERHS (020, 046, 082, the latter two improving on all three measures). On the whole the evidence suggests that there is a clear sense of impoverishment as compared with the previous generation.
3.5.1 Reasons for improvement and decline in
Standard of Living
Considering the Standard of Living Graphs among the rich and poor, 3 very rich, 2 rich, and 1 poor mentioned improvement. The main reasons for improvement among the very rich households were said to be access to be good prices for irrigated crops (mainly onions and fruit) especially in the last three years, and involvement in trade using camels. Reasons for improvement among the rich they included investing money from weaving to finance sharecropping, good parental endowments and involvement in the Kebele. Only one poor household showed a slight improvement owing to good harvests, involvement in daily labour, the cow his new wife brought and her income from cotton spinning. Another poor man had come from a family that had to send him to become a servant. He gradually earned the right to leave and his master helped him establish himself.
Reasons for decline among the very rich seem to be the ageing of the household head and loss of labour with children leaving. Among the rich reasons included livestock loss due to drought and disease, labour shortage with children leaving, decreasing agricultural output and increasing number of household members, production shocks with crop and livestock losses and shortage of labour with a daughter leaving and the wife dying. Among the poor the main reason for decline seems to be the ageing and disabilities of the household head and death of the household head followed by household disintegration, and loss of labour with adult children departing. Among the very poor the main reasons seem to be ageing, disability and limited support with children leaving or unhelpful. Other reasons include loss of assets in a fire, theft of livestock, land returned due to disability and the remaining plot given to sharecroppers and subject to tax, and for female headed households death of the household head, giving out land to sharecroppers and land confiscation.
3.6. Global Happiness
Score
Out of the 34 households the majority, 19 cases, had a GHS of 2 or 'fairly happy'; 9 had a GHS of 3 or 'not so happy' and only 6 had a GHS of 1 or 'very happy'. If we consider the households at the two extremes, among the six households with a GHS of 1 only one was very rich (and considered himself to be 'among the richest), two were rich (but considered themselves to be 'about average), two were middle wealth (one of whom considered himself to be average and the other 'the poorest', and only one was poor. Let us consider the cases of the poor and middle households.
The poor case (056) is a
household composed of a young man of 20 his wife of 16 and three sisters. His
mother had died leaving him to look after the household and he was recently
married. The household had 1.5 hectares of land, none of which was irrigated.
They had 1 bull, 1 goat, and 1 chicken. They were in the 3rd asset
quintile ranked 16 out of 34. The household was ranked as 'among the poorest',
and 'a little worse off' than a year ago, but 'much worse off' than 5 years
ago. The household head views himself as poorer than his father. The
contentment graph shows a drop to rock
bottom at 1. The head mentioned that their conversion to Islam caused problems
with their family and that he is still not happy about it. The GHS
is somewhat surprising, especially since the interview with the
household revealed that the younger children were still suffering from the loss
of their mother, and the standard of living and contentment graphs coincide in
revealing sharp declines. However, it may be that the head's recent marriage
was a factor that contributed to his happiness, despite his assessment of being
much worse off than five years ago and a little worse off than a year ago.
One of the middle wealth cases (009) is a household with 5 members: The male head aged 68, his wife, aged 65, two sons aged 15 and 13 and a grand-daughter aged 15. The first wife left in 2003 and the new wife came the same year and the household head expressed satisfaction with the new wife who is treating the children well. The household had 0.81 hectare of land, less than half of which was irrigated. The household had 1 ox, 1 bull, 1 cow, 1 heifer, 20 chickens and four bee-hives. It was in the 3rd quintile in terms of assets ranked 19th out of 34. The RANS responses put the household in the 'about average' category, and at 'the same' as 5 years ago, and also 'the same' as one year ago. The household head suggests he is 'poorer' than his father. The Standard of living timeline suggests that there have been improvement in 1998 with better harvests, and especially from 2002, and that the standard of living has been improving in the past three years owing to income from cash crops. However, the SLG does not show a very significant improvement. The household was placed at 'average' in imperial and Derg periods, with a slight drop within that category in the early EPRDF period and a slight rise within that category becoming borderline to the next category 'richer than most' in the past three years. T