Discussion Paper 65: REDDY, Ratna V.: "Wealth Ranking" in Socio-economic Research: Substitute or Complement?

Diskussionsschriften der Forschungsstelle für Internationale Wirtschafts- und Agrarentwicklung eV (FIA), Nr. 65, Heidelberg 1997

 
____________________________________________________________________

 

Abstract: This paper is an attempt to compare wealth ranking approach with other standard approaches of income analysis followed in socioeconomic research in order to test the validity of wealth ranking on one hand and its relevance in di fferent economic situations. This paper, while validating the use of wealth ranking, suggests that it can generate differential and perhaps better results in some specific contexts. Wealth ranking could be considered as an alternative in the context of ru ral communities where nonagricultural incomes dominate. However, its general application is constrained by theoretical as well as practical problems associated with it. Unless these limitations are addressed effectively, it is unlikely that wealth ranking can be considered as a substitute for standard approaches. This study, based on the comparative analysis of two different situations (economic), emphasizes its use as a complementary approach to the standard methods of income analysis in socioeconomic re search.

I Introduction

Participatory Rural Appraisal (PRA) methods are considered, by certain academics, as a better alternative to traditional survey / questionnaire methods in socioeconomic research. Of late these methods have not only gained prominence but also propagated as an alternative paradigm in social science research (Chambers, 1994a; b and c). In fact, some of the funding agencies are promoting the use of PRA in place of survey methods. The main advantage of PRA methods is its cost effectiveness and si mplicity. PRA is an alternative mode of information gathering with a stress on qualitative information rather than on quantitative information. According to Chambers (1992) PRA is intended to enable local people to conduct their own analysis and often to plan and take action. PRA has a menu of methods to elicit information from rural people. These methods range from secondary sources to semi-structured interviews and simple questionnaires. PRA though boasts of twenty-seven methods, most of them are being used in the standard approaches of socioeconomic research in one form or other, of course under different names or even at informal level (for a critical review of PRA see Reddy, 1998). However, PRA differs with regard to participatory mapping and diagram ing, well being or wealth ranking and analyzing. Of these wealth ranking is the most widely used method and has parallels in the standard survey methods.

In the standard socioeconomic analysis size of holding (farm size) or household income or consumption is used to categorize households into different wealth groups. These groupings are used to identify the most needy households for the purpose of a tar get group oriented poverty alleviation programme or to examine the variations in different socioeconomic attributes across income groups. This information is collected directly from the head of the household (or respondent) with the help of a questionnair e. The reliability of such data is often questioned on the grounds that, there exists differences and contradictions between observed or experienced realities and the reality generated from survey data (Jodha, 1986; Bardhan and Rudra, 1978). These differe nces further get accentuated in situations where cultural and traditional value systems predominate. For instance, households belonging to higher social strata often do not like to project an economically poor status, especially to an outsider. Though som e households still maintain their life styles (reflecting in the consumption information), they may be increasingly getting into debt which the household is unlikely to reveal. But the fellow villagers are aware of the reality. Similarly, the vice-versa w ould be true in the case of households who are not willing to reveal their latent assets and other sources of income such as derived from money lending.

Wealth ranking, on the other hand, is carried out by groups of villagers, i.e., villagers rank each and every household according to their wealth and well-being. It is argued that wealth ranking provides more accurate and realistic categorization of ho useholds by their wealth compared to the standard questionnaire method as the villagers take various aspects of the household into consideration while ranking. Apart from the cost effectiveness of the method, the arguments against the standard approach an d in favor of wealth ranking sound logical and convincing. But one is not sure how good is this logic on empirical grounds. Empirical testing of the differences between the two methods is necessary before discarding or adopting either of the methods. For, the acceptability of PRA methods in the main stream research is still dubious due to; i) data generated through PRA may not stand the statistical rigor, and ii) how far PRA can provide more authentic or different analytical insights compared to standard survey methods. Besides, the purpose to which the data, thus generated, is used for is also important in choosing the methods. This is important not only to determine the suitability of the data set for statistical testing but also to determine the cost e ffectiveness of the PRA methods.

In this paper an attempt is made to compare wealth ranking with income analysis followed in the standard survey methods in order to examine the differences or similarities between them (ii above). Hither to, income analysis in the main stream socioecon omic research is based either on size of land holding (farm size), or on household income or on expenditure. Of late some of the studies using PRA techniques are following wealth ranking approach. The validity of these two methods has confined to theoreti cal debate only as the studies so far adopted either of the two approaches rather than using both of them and testing the difference between them. In a recent study Adams et.al (1997) validated the use of wealth ranking by observing variations in socioeco nomic and demographic variables across wealth ranking groups. While this study supports the use of wealth ranking for socioeconomic stratification, it does not provide any insights regarding how wealth ranking compares with the standard household income a nalysis. That is whether or not wealth ranking can provide different analytical insights compared to the standard analysis. Proving the superiority of wealth ranking would mean questioning the validity of the conclusions hither to drawn from a number of s tudies adopting the standard sizeclass wise approach for income analysis. This aspect is examined here using the data generated through wealth ranking as well as standard survey technique in an intensive study of two villages in Rajasthan state of western India. The data are drawn from a larger study "User Valuation of Renewable Natural Resource: A Study of Arid Zone," where participatory and survey methods were used as complements. Wealth ranking and questionnaire were used for all the households in both the villages (census). The main objectives of this paper are: i) to examine the relation between wealth ranking and size of holding (farm size) on one hand and wealth ranking and household income on the other, and ii) to examine the variations in land us e patterns and demographic aspects across households when stratified according to wealth ranking and farm size.

II Wealth Ranking: The Approach

Wealth ranking or well-being ranking is an indirect approach to obtain information on households’ economic status. Usually, socioeconomic status of the household is obtained through a questionnaire where the household is asked directly regar ding its socio-demographic characters, economic assets and income / expenditure. However, data collected through a questionnaire may not reflect the true economic conditions of the household due to the following reasons; i) asset position (land, buildings , etc.) may not provide continuous flow of income under all circumstances, ii) there are other sources of income such as income flows from migrant labor, interest payments from money lending, etc., which are not covered under assets, and iii) households o ften tend to underestimate / overestimate their income / expenditure due to various reasons. On the other hand, wealth ranking is carried out by the villagers based on their own perceptions of the particular household. Villagers while identifying the econ omic status of a household consider beyond its asset position. Moreover, since the issue is discussed within a group of individuals, various dimensions of wealth will be taken in to account and hence improves the objectivity of identification.

Wealth ranking may be carried out by different groups of people within the village. These groups may belong to different communities or genders. According to the perceptions of these groups’ ranking may differ. For instance, in one of the wealth rankin g exercises in Uganda when men and women were asked separately to rank the households, it was observed that while men concentrated on material wealth women gave more weighage to social circumstances (Seeley, et. al., 1996). However, comprehending these di fferences will be a difficult task if one is interested in studying the economic status of the households in a village. That is depending on the nature of the study due weighage should be given to material and social wealth.

As far as socioeconomic research is concerned, the objective of income analysis is to examine the variations in some key variables such as resource use patterns across income groups. Wealth ranking exercise is carried out here with the same objective. As a part of the major study, mentioned earlier, two villages were selected for carrying out an intensive study of resource use pattern and valuation. The demographic aspects along with socioeconomic aspects of the selected villages are presented in Table 1. Information on socioeconomic, demographic and resource use patterns were obtained from all the households in both the sample villages. In order to get the economic status of the households, data pertaining to land assets (farm size), agricultural and nonagricultural incomes were elicited. These variables form a base for grouping the households according to their economic status.

As a second base we have carried out wealth ranking exercise by two groups of villagers in each village. Since some of the households live in farm houses (dhanis) outside the village groups were formed in such a way to cover both dhanis a nd main villages. In both the villages the participants were very enthusiastic about the exercise and participated eagerly unlike in the case of questionnaire filling. In both the villages it was a new experience for the participants. There are no non-gov ernmental organizations (NGOs) working in the villages. The rapport with the villagers is built by the research team over a period of one year involving intensive study of the villages. The groups at dhanis were relatively smaller given the limited number of households living there. About eight people participated in the wealth ranking exercise at the dhanis (one from each village) while about 15 people participated in the main villages. Before starting, the purpose of the exercise was discu ssed with the group and also regarding what the participants are expected to do. For the purpose of the exercise, cards were prepared for each household with the name of the head of the household is written on the card. The team members have read out each household’s name for the group and the group was asked to rank the economic status of the household. No clues regarding the indicators (land or other assets) were given to the group. The group was requested to give the lowest rank (one) to the poorest ho usehold and highest rank to the richest household in the village. Accordingly all the households were ranked between discussions among group members. Same ranks were given to 2-3 households in a few cases. After the exercise was over the participants were requested to explain the criteria they have used for ranking the households. The main criteria / indicators they have taken into account are i) ancestral property including land and gold, ii) involvement in money lending, iii) sources of nonagricultural income such as government service or business, and iv) remittances from outside by family members.

Table 1: Socioeconomic and Demographic Aspects of the Sample Villages

Item

Jheengar Bari

Khori Brahman

I) No. Households

83

118

ii) Total Population

503

701

iii) Amenities

 

 

a) Drinking Water

YES

YES

b) Medical

NO

NO

c) Education

Primary

Primary

d) Transport

NO

YES

e) Electricity

YES

YES

iv) % of Households belonging to the Category

 

 

a) Landless Laborers

11

0

b) Marginal Farmers

10

10

c) Small Farmers

17

14

d) Semi-medium Size Farmers

28

31

d) Medium Size Farmers

30

24

e) Large Farmers

4

10

v) % of Households depending on
Agriculture [Major Source of Income]

58

40

vi) % of Area Irrigated [Net]

20

03

vii) Proportion Nonagricultural Income*
[Average]

57

87

Note: * Major nonfarm activities include nonfarm labor and remittances from out side.

Before examining the relationships between economic variables and wealth ranking it would be appropriate to present the important economic features of the sample villages. For, economic status of the villages has a bearing on the relationships a nd hence facilitates their explanations. The main occupation in both the villages is not agriculture in terms of income source though a majority of the households (57 percent) depend on agriculture in Jheengar Bari (Table 1). The dependency on nonagricult ural activities, in terms of number and income derived, is much higher in Khori Brahmanan compared to Jheengar Bari village. Though there are no landless laborers in Khori Brahmanan the dependency on wage employment is more compared to the other village. This may be attributed to lack of irrigation facilities in the Khori Brahmanan village. Besides, in Khori Brahmanan number of people migrate to other states in pursuance of business and wage employment and hence remittances from outside account for a majo r share in the household income. Within the farming community, both the villages have predominantly medium size farmers, though the average size of land holding is higher in Khori Brahmanan. The lower size of holding in Jheengar Bari is due to the availab ility of irrigation and the resultant intensive farming practices.

III Wealth Ranking vis-a-vis Economic Variables

As reflected in the economic features, the sample villages differ from each other in terms of economic conditions. The two important aspects that are relevant for the relationships between wealth ranking and other standard economic variables are: i ) the dependency on nonagricultural income is substantially higher in one of the villages. This would make farm size a less important economic variable and hence one usually depends on household income for measuring the economic status in such situations; ii) relevance of farm size is also questionable when land is not a major constraint which is the case in fragile resource regions like Rajasthan. This coupled with the biases in reporting of household income prompts one to rely on alternative indicators like wealth ranking. In this section, we examine how wealth ranking compares with standard economic variables like farm size, agricultural, nonagricultural and total household incomes.

We have estimated the simple correlation coefficients between wealth rank and other economic variables in order to compare them. As mentioned earlier wealth ranking exercise was carried out with two groups of people in both the villages. On the whole 8 3 households in Jheengar Bari and 100 households in Khori Brahmanan were ranked. The difference between the ranked and actual households in Khori Brahmanan is due to the reason that some of the households temporarily migrate to other places in search of e mployment or business though they own land and other assets (houses) in the village. Accordingly we have dropped the nonresident households from the main list and matched the remaining. The consistency in ranking between the two groups is very high in bot h the villages. The rank correlations are as high as +0.99 and significant at less than 1 percent level. This indicates that there are no inconsistencies in the ranking and the criteria adopted by the two groups of participants in the exercise.

Table 2: Relationship Between Wealth Ranking and Other Income
Variables in the Sample Villages

 

Income Variables

Jheengar Bari

Khori Brahmanan

Rank I

Rank II

Rank I

Rank II

1. Agricultural Income

0.51*
(5.34)

0.51*
(5.33)

0.20***
(2.02)

0.22**
(2.23)

2. Nonagricultural Income

0.16
(1.43)

0.16
(1.46)

0.42*
(4.57)

0.41*
(4.52)

3. Total Income

0.38*
(3.70)

0.38*
(3.71)

0.47*
(4.65)

0.47*
(5.27)

4. Own Area [Farm Size]

0.67*
(8.16)

0.68*
(8.25)

0.34*
(3.50)

0.36*
(3.81)

Note: Figures in brackets are 't' Values. *, ** and *** indicate levels of significance at 1, 5
and 10 percent respectively.

The simple correlation coefficients are estimated between wealth ranks on one hand and agricultural income, nonagricultural income, total income and farm size on the other. All the coefficients carry positive signs and all of them, except nonagr icultural income in Jheengar Bari, are significant (Table 2). The significant positive correlation indicates that there is one to one correspondence between wealth ranking and the standard economic variables. In other words, the economic status of the hou sehold reflected in the standard income / economic variables is the same as that of what the villagers perceive as economic / wealth status. The relationships are more prominent, in terms of the magnitude of the coefficients and their level of significanc e (‘t’ values), in the case of dominant income sources. In Jheengar Bari, where agriculture is the major activity, villagers have given high priority to farm size and agricultural income compared to other variables. Whereas in Khori Brahmanan nonagricultu ral as well as total income of the household are given high priority. Overall, the dominant activity seems to be reflecting in the wealth ranks. The positive and significant relationship between farm size and wealth ranking even in Khori Brahmanan may be due to the reason that those households with more land assets are also having nonagricultural assets like gold and income flows from business or money lending. The stronger relationship between wealth ranking and farm size (and relatively weak relationshi p between wealth ranking and total income) in Jheengar Bari and between wealth ranking and total income (and relatively weak relationship between wealth ranking and farm size) in Khori Brahmanan indicates wealth ranking is more relevant in communities wit h a larger share of nonagricultural income in the household income. For, farm size does not reflect the true picture in such conditions and reporting of household incomes is less accurate than area owned (farm size).

These results can be interpreted in two ways. Firstly, wealth ranking is comparable with other standard economic variables as far as generating information relating to households economic status is concerned and hence the use of wealth ranking is valid ated. Secondly, since the information generated by the two methods does not seem to differ much, it is difficult to establish the superiority of one method over the other. However, wealth ranking is likely to provide more accurate information in the conte xt of communities where land is not a dominant source of income, especially in the light of less reliable reporting of household income. These aspects coupled with its cost effectiveness, wealth ranking seems to score over the standard methods. At this ju ncture it is pertinent to examine how an analysis based on wealth ranking compares with that of a standard procedure (farm size wise analysis) at a disaggregated level. This would help us in examining not only the variations in the socioeconomic variables between the two methods, but also to observe the differences in the analytical insights one can draw from using the two methods. For, the differences in the analysis, if any, between the two methods would get accentuated at the disaggregate level. This a spect is taken up in the following section where wealth ranking wise analysis is compared with size class wise analysis with regard to the distribution of some important variables.

III Wealth Rank Wise Vis-a-vis Farm Size Wise Analysis

It is a common practice in socioeconomic research concerned with rural communities to group the data according to land holding classes. For the purpose of comparison we have grouped the households by their farm size status as well as wealth rank st atus. For wealth rank wise categorization, all the households are arranged in ascending order of their wealth ranks and then divided into five equal groups. These groups may be treated as proxies to the size class wise groupings, i.e., marginal, small, se mi-medium, medium and large. Though these two groupings may not be strictly comparable, they do reflect the situation in the respective categories. That is, while marginal farmers’ category reflects the situation in the lowest category of the farming comm unity, the first (lowest) group in the wealth rank represents the situation in the poorest category of the households. Apart from this there is no room for bias as we have taken all the households into account in both the villages. The comparative analysi s is carried out for land use patterns, cropping patterns and family size. In the case of land use and cropping patterns analyses we have dropped the observations pertaining to the land less category from both the approaches. There are nine households in the landless category in the Jheengar Bari village.

As far as land use patterns are concerned one of the important aspects often addressed is the relationship between farm size and intensity of land use. It is widely observed, in the Indian context, that lower size class farmers use land more intensivel y while large farmers use it extensively (Reddy, 1991). This is often attributed to higher proportion of area under irrigation with the small farmers. However, this may hold good only in the case of public irrigation while the reverse is observed in the c ase of private (well) irrigation (Vidyasagar and Reddy, 1996). Here, these hypotheses are examined on the basis of land holding size and wealth ranking groupings. The variables considered are proportion of area under grazing and pasture lands, fallow land s and irrigation. Our intention here is not to test the hypothesis as such, but to observe the differences, if any, in the analysis between wealth rank wise grouping and farm size wise grouping of the households.

Table 3: Land Use Pattern Across Wealth Rank and Farm Size Classes in the Sample Villages

Wealth Rank
(Farm Size)

No. of households

Average Size of holding (ha.)

% of Pasture & Grazing Land

% of Fallow Land

%of Area Irrigated

Jheengar Bari

I (Marginal)

15 (08)

1.2 (00.5)

1.8 (00.0)

01.8 (00.0)

13.4 (00.0)

II (Small)

15 (12)

3.1 (01.4)

1.9 (01.7)

04.0 (05.6)

29.4 (20.5)

III (Semi-medium)

15 (25)

5.1 (02.7)

0.4 (01.1)

02.5 (02.7)

32.8 (32.1)

IV (Medium)

15 (25)

5.4 (07.1)

1.0 (01.0)

02.6 (02.5)

33.6 (36.9)

V (Large)

14 (04)

6.7 (12.1)

1.2 (01.1)

03.1 (03.2)

44.0 (40.5)

ALL

74 (74)

4.3 (04.3)

1.1 (01.1)

02.9 (02.9)

34.8 (34.8)

Khori Brahmanan

I (Marginal)

20 (01)

2.9 (00.3)

2.3 (00.0)

12.9 (00.0)

09.6 (00.0)

II (Small)

20 (28)

3.2 (01.6)

2.5 (00.0)

10.1 (19.0)

00.0 (05.7)

III (Semi-medium)

20 (32)

3.9 (03.0)

1.5 (02.4)

15.9 (12.1)

08.0 (04.4)

IV (Medium)

20 (33)

3.4 (06.1)

1.2 (02.8)

15.9 (12.6)

08.0 (04.4)

V (Large)

ALL

20 (06)

100 (100)

9.8 (22.0)

4.7 (04.7)

5.7 (06.5)

3.5 (03.5)

10.3 (10.0)

12.3 (12.3)

07.9 (11.9)

06.4 (06.4)

Note: Figures in brackets indicate farm size wise information. Wealth rank I indicates the poorest category of households and V indicates the richest category of households. Marginal= Up to 1 hectare, Small= 1-2 hectares, Semi-medium=2-4 h ectares, Medium= 4-10 hectares, Large= Above 10 hectares.

It may be noted that the average size of holding does not match between wealth ranking and farm size wise groupings (Table 3). The lower (poorer) category households are having larger farm size when grouped according to their wealth ranking than when grouped according to their land holding. And the vice-versa is true in the case of large farmers. This is more so in the case of Khori Brahmanan where nonagricultural incomes dominate. In fact, in this village medium size farmers are owning less lan d, on average, than semi-medium size farmers. These discrepancies exist even when wealth ranked households are grouped (in numbers) according to farm size wise groupings (Appendix, Table 1). This indicates that land assets are not the prime concern while ranking the households and hence may not reflect the well-being of the household. These differences seem to be present even in the land use pattern though the trends across size classes on one hand and wealth rankings on the other are some what similar, e specially in Jheengar Bari (Appendix, Table 2). Due to the mis match between wealth ranking and land ownership it is observed that poor households (as per wealth ranking) having irrigation facilities and have a portion of their land under pasture & gr azing and fallow categories. The relationship between farm size and proportion of area under irrigation is a smooth and positive one, while it is not smooth when the households are grouped according to their wealth ranking. This is more so in the case of less irrigated village.

It may be noted that the differences between wealth ranking groupings and farm size groupings are relatively less in the agriculturally dominant village (Jheengar Bari) compared to the village where nonagricultural incomes play a major role. This indic ates that wealth ranking might provide differential results in the case of communities where nonagricultural incomes dominate. Therefore, using farm size to differentiate economic classes in all situations may result in erroneous conclusions. In the prese nt context land ownership and access to irrigation, which are the standard variables in determining the economic status of the household, do not seem to reflect the well-being of the household, as perceived by the villagers. For, the poorest category of t he households are owning more land and having irrigation facilities in the case of wealth ranking while farm size wise categorization shows the contrary. Discrepancies are observed even in the case of area under fallow. As the access to irrigation determi nes the cropping pattern to a large extent, differences between the two approaches are also observed in the case of cropping patterns across wealth groups (Table 4). Here we have grouped the wealth ranked households according to the farm size wise categor ization as the differences between equal division and arbitrary division are marginal. It may be noted that the discrepancies are more in the case of less irrigated (Khori Brahmanan) village.

The demographic aspects of the household are also analyzed in terms of family size. Interestingly, both the approaches have brought out similar results (Table 5). It is generally assumed that poor households have larger family size as compared to the r ich. However, this hypothesis has been questioned by some of the recent empirical studies in India and else where ( Vyas, 1991; Reddy, 1996; Adams, et. al, 1997). These studies are, in fact, in support of a reverse hypothesis, i.e., large farmers have big ger family size compared to their counter parts. The present analysis also supports this reverse hypothesis for both the approaches. These results vindicate some of the recent observations in the context of Rajasthan in this regard (Reddy, 1996). Therefor e, these results validate the use of wealth ranking approach.

 

 

Table 4: Cropping Pattern Across Wealth Groups in the Sample Villages
(Major Crops)

Wealth Rank

Kharif (% Area)

Rabi (% Area Under)

(Farm size)

Bajra

Pulses

Wheat

Gram

Mustered

Onion

Jheengar Bari

I (Marginal)

72 (88)

28 (12)

00 (00)

00 (00)

00 (00)

00 (00)

II (Small)

47 (51)

40 (31)

07 (10)

02 (00)

02 (03)

02 (04)

III (Semi-medium)

42 (44)

33 (21)

11 (09)

02 (00)

02 (06)

07 (06)

IV (Medium)

34 (37)

34 (14)

12 (10)

06 (05)

06 (05)

03 (05)

V (Large)

46 (31)

27 (20)

13 (14)

00 (04)

02 (05)

01 (03)

ALL

39 (39)

34 (34)

11 (11)

04 (04)

04 (04)

04 (04)

Khori Brahmanan

I (Marginal)

100 (100)

00 (00)

00 (00)

--

--

--

II (Small)

64 (80)

31 (20)

04 (00)

--

--

--

III (Semi-medium)

60 (66)

34 (27)

04 (07)

--

--

--

IV (Medium)

51 (54)

41 (42)

04 (03)

--

--

--

V (Large)

46 (41)

52 (47)

02 (06)

--

--

--

ALL

56 (56)

38 (38)

04 (04)

--

--

--

Note: Same as in Table 3.

 

Table 5: Average Family Size Across Wealth Groups in the Sample Villages

Wealth Rank

Average Family Size

(Farm size)

Jheengar Bari

Khori Brahmanan

 

A

B

A

B

I (Land Less)

5

4 (3)

--

-- (--)

II (Marginal)

6

6 (7)

8

2 (3)

III (Small)

6

6 (7)

7

8 (6)

IV (Semi-medium)

7

7 (6)

9

8 (7)

V (Medium)

9

8 (9)

8

9 (10)

VI (Large)

8

9 (9)

10

10 (15)

Note: Same as in Table 3.

A= When wealth ranks are grouped in equal numbers.

B= When wealth ranks are grouped according to farm size wise groupings.

V Conclusions

The preceding analysis is very indicative of the validity of wealth ranking vis-a-vis the standard approach in social sciences research. It suggests that wealth ranking can provide differential and perhaps better results when compared to the standa rd farm size wise analysis in the context of rural communities where nonagricultural incomes have a major share in the total household income. In this regard wealth ranking could be considered as an alternative to the standard approach. Where as in the do minantly agricultural communities the inferences may not differ much between these two approaches. In such situations wealth ranking can score over the standard approaches in terms of its cost effectiveness. However, the cost effectiveness depends on the nature and purpose of the study one is under taking. It would be cost effective as long as the study objective is limited to identifying policy needs and development planning such as identifying the poorest of the rural households. When the objective is p olicy research where one is concerned more with quantification and statistical rigor and robustness of the analysis, wealth ranking can only be used as a complementary tool to the standard questionnaire approaches rather than a substitute. For, it is ofte n observed that participatory methods are constrained by the fact that they cannot go beyond identifying policy needs ( Greely and Rowshan, 1995).

While it is clear that wealth ranking could be considered as a substitute to the standard approach in nonagricultural communities in a study of limited objectives such as identifying the policy needs, its application in general is constrained by theore tical as well as practical problems associated with it. Some of the important limitations are:

Unless these limitations are addressed effectively, it is unlikely that wealth ranking can be considered as a substitute for the standard approaches. This calls for further research on comparing both the approaches in a variety of situations. The prese nt study while focusing on two different situations validates the use of wealth ranking and emphasizes its use as a complementary approach to the standard methods of income analysis in social sciences research. Wealth ranking though provides valuable info rmation, its interpretation needs careful analysis rather than taking it on its face value. Wealth raking would be more valuable when used as a compliment to other methods of economic analysis instead of using it as an alternative.

References

Adams A M et. al (1997), "Socioeconomic Stratification by Wealth Ranking: Is it Valid?, World development, Vol. 25, No. 7, July.

Bardhan, P and Ashok Rudra (1978), "Inter Linkages of Land-Labour and Credit Relations - An Analysis of Village Survey Data in East India", Economic and Political Weekly, Annual Number, February.

Chambers, Robert (1992), Rural Appraisal: Rapid, Relaxed and Participatory , Discussion Paper 311, Institute of Development Studies, Sussex, U. K.

Chambers, Robert (1994a), "The Origins and Practice of Participatory Rural Appraisal", World Development, Vol. 22, No. 8.

Chambers, Robert (1994b), "Participatory Rural Appraisal [PRA]: Analysis of Experience", World Development, Vol. 22, No. 9.

Chambers, Robert (1994c), "Participatory Rural Appraisal [PRA]: Challenges, Potentials and Paradigm", World Development, Vol. 22, No. 10.

Greely, Martin and Rabeya Rowshan (1995), " Assessing Poverty and Well Being in Rural Bangladesh: Motives, Measures and Methods," Paper presented at the Workshop on The Village in Asia Revisited, Center for Development Studies Trivandrum, 26-28, January.

Jodha N. S (1986), "Social Science Research on Rural Change: Some Gaps (A Footnote to the Debate on Rural Poverty)," International Crops Research Institute for the Semiarid Tropics (ICRISAT), Hyderabad.

Maxwell, Simon and Clande Bart (1995), "Beyond Ranking: Exploring Relative Preferences in P/RRA," PLA Notes, No. 22, IIED, London, February.

Reddy, V. Ratna (1991), "Under Utilization of Land in Andhra Pradesh: Extent and Determinants", Indian Journal of Agricultural Economics , Vol. 46, No. 4, October- December.

Reddy, V. Ratna (1998), "User Valuation of Renewable Natural Resources: Some Methodological Issues", Quarterly Journal of International Agriculture , Vol. 37, No. 1

Reddy, V. Ratna (1996), A Study of Willingness and Ability to Pay for Water, Project Report, Institute of Development Studies, Jaipur.

Richards, Paul (1995), "Participatory Rural Appraisal: A Quick and Dirty Critique", PLA Notes, No. 24, IIED, London, October.

Seely, Janet, et. al., (1996), " Revealing Gender DifferencesThrough Well-being Ranking in Uganda", PLA Notes, No. 25, London, February.

Vidyasagar and V. Ratna Reddy (1996), "Measuring Inequality in the Distribution of Irrigation: Some Methodological Issues with Reference to Irrigation Development in India", Working Paper, Institute of Development Studies, Jaipur, India.

Vyas, V. S (1991), "Agrarian Structure, Environmental Concerns and Rural Poverty", Elmhirst Memorial Lecture, XXI International Conference of Agricultural Economists, Tokyo, Japan, 22 August.

 

Appendix

Table 1: Land Use Pattern Across Wealth Ranks When Number of Observations are Equivalent to Farm Size Grouping

Wealth Rank

No. of HH

Average size of holding (ha)

% of pasture and grazing land

% of fallow land

% of area irrigated

Jheengar Bari

 

 

 

 

 

I

8

0.71

0.0

0.0

0.0

II

12

2.42

2.76

4.83

13.79

III

25

4.18

0.69

2.53

35.10

IV

25

5.65

1.11

2.95

41.86

V

4

7.97

1.00

2.51

27.23

Khori Brahmanan

 

 

 

 

 

I

1

1.30

0.0

15.00

0.0

II

28

3.62

2.48

11.50

5.60

III

32

3.24

1.65

14.88

3.14

IV

33

6.76

4.25

11.31

8.00

V

6

7.17

6.04

12,45

7.55

 

Table 2: Cropping Pattern Across Wealth Ranks when Number of Observations are Equivalent to Farm Size Grouping

Wealth Rank

Kharif (% area under)

Rabi (% area under)

Bajra

Pulses

Wheat

Gram

Mustard

Onion

Jheengar Bari

 

 

 

 

 

 

I

72

28

0.0

0.0

0.0

0.0

II

47

40

6.6

1.6

1.6

2.0

III

42

33

10.7

2.4

1.9

6.8

IV

34

34

12.4

5.5

6.0

3.0

V

46

27

13.4

0.0

2.0

1.0

Khori Brahmanan

 

 

 

 

 

 

I

100.0

0.

0.0

-

-

-

II

64.4

31.0

4.0

-

-

-

III

60.4

34.0

4.3

-

-

-

IV

51.2

41.0

4.2

-

-

-

V

45.6

52.0

1.7

-

-

-