You are on page 1of 14


Gautam Pingle
This paper attempts to compare the relative economic status of rural Hindu households with
that of rural Muslim households over the decade 1994-2004. It relies for base line estimates on the
rural surveys conducted in 1994 by the National Council for Applied Economic Research. It also
uses the National Sample Survey Organisation’s 61st Round’s 2004 rural data as extracted and
published in the Sachar Report. It further examines data on Hindu-Muslim income differentials
identified as a result of the independent surveys of 42 Minority Concentration Districts (MCDs).
The findings do not substantiate the idea of fixed and firm Muslim household income/expenditure differences over time compared to the Hindu majority. On the other hand, they indicate that
a relatively faster Muslim progress over the decade 1994-2004 has led to closing of the initial gap.
This is especially heartening as it relates to the period of liberalisation and the transition to a market
economy, which has disrupted many traditional patterns of rural livelihoods especially of artisans.

The status of minorities in India is always
important. Yet keeping track of their progress is
no easy task.1 The "High Level Committee Report
on Social, Economic and Educational Status of
the Muslim Community of India" [Sachar, 2006]
(hereinafter referred to as Sachar or Sachar
Report) remains the first comprehensive attempt
at this task. Not only did it present data on
Muslims; it also gave comparative data on the
Hindu majority as also on Other Minorities, (i.e.,
non-Muslim religious minorities). It generated
widespread interest both in the academic and
political world and in civil society. It also stimulated policy initiatives by the Union and State

This paper attempts to compare the relative
economic status of Hindu and Muslim rural
households over the decade 1994-2004. For the
1994 base line, we use data generated by the rural
surveys conducted by the National Council for
Applied Economic Research (NCAER). For
2004, we depend on the data, published in the
Sachar Report, which was extracted from the 61st
Round3 of the National Sample Survey Organisation (NSSO). We have perforce to deal with the
data issues posed by the Sachar Report. The paper
also examines other data on Hindu-Muslim
income differentials identified as a result of surveys of Minority Concentration Districts
The paper concentrates entirely on the
income/expenditure of households, as this is the
resultant outcome of household resources,
household preferences, individual educational
qualifications and general job opportunities.
Income of a household largely depends on its size,
the age and gender profile of its members and the
degree to which their members are gainfully
employed. We will deal with these issues in due

The Sachar Report has achieved legitimacy as
a political document on Muslim deprivation
despite some major data and methodological
issues relating to its economic sections; issues
which call for caution while making policy.
Considerable controversy has arisen after the
publication of the Sachar Report leading to a
reappraisal by policy makers.2 The Report should
be considered only as the precursor to establishing
a regular series of credible economic databases
This paper does not address elite issues such
on which more robust and objective analysis can as jobs in civil service, armed forces and police be made and effective and robust policy initiatives or even in the representation in the legislatures
and judiciary and jobs in associated offices. It

Gautam Pingle is Director, Centre for Public Policy and Governance, Administrative Staff College of India, Bella Vista,
Raj Bhavan Road, Hyderabad - 500082. E-mail:

Muslims in all the states have recorded lower than state average consumption levels." [Sachar. This is largely because both the NCAER and NSSO samples were drawn to represent the entire state population and any . We are also conscious of the fact that the sampling frame and methodology of both surveys differ and the data are not exactly comparable and are estimates to be treated with caution.6) underlines the inequity existing in these areas. relative to the other SRCs. 500 expenditure class. the proportion of H-General and ‘all others’ in this class is much lower. In fact in West Bengal. Sachar commented thus: This paper attempts to compare the data presented in Sachar with a study conducted ten years earlier in order to examine any changes over the period. the share of such households among Muslims. In urban areas in almost all the states where Muslim proportion is high. One of the Terms of Reference of the Sachar Committee (regarding Muslims) was to determine: "What are their asset bases and income levels relative to other groups across various States and Regions? [Sachar. the condition of Muslims is relatively better in rural areas. Gujarat. however.2 and 8. to consider the household incomes/expenditures of the mass of Hindus and Muslims who live and work in the rural areas where income generating opportunities are not as abundant as in urban areas and where traditional and religious constraints can come in the way of economic progress of households. 154] (Emphasis added).that is. The comparison over the decade is limited to the rural sector and to nine states.3). pertains only to the rural sector and gives the Hindu and Muslim average annual income data. The 1994 NCAER survey. Andhra Pradesh. rather than their levels of household expenditure and concluded that there was inequity." [Sachar. 2006. Fig. 500 to Rs. While there is a substantial proportion of households in all SRCs in the Rs.6 present[s] estimates of MPCEs for selected states for urban areas for different areas. such as Hindu-General (H-General) (elsewhere also labeled "Other" and implying Upper Caste) with No attempt was made by Sachar to reconcile its two conclusions regarding rural expenditure Muslims is that: differentials. SCs/STs and OBCs households are located in the below Rs. OBCs and SCs/STs is lower.86 JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY seeks. DATA ISSUES Since the Sachar and NCAER Hindu-Muslim data available is of a collateral nature . However. The proportion of households from these three SRCs with expenditure levels above Rs. p. 8. 2006. 2006. on the other hand. p. the MPCE of Muslims is substantially below that of other SRCs except SCs/STs. The conclusion that Sachar came to by comparing Socio-Religious Categories (SRCs). 155] (Emphasis added). 8. 1000 expenditure bracket. Excepting Assam. it would seem that this aspect of income has not received the attention it deserves. vi]. although the MPCE level itself is much lower than that in the urban areas. 1000 is also very low. Rajasthan and Madhya Pradesh they have levels lower than even the SCs/STs. when it came to dealing with MPCE levels for different SRCs. We are limited to just nine states for which comparable data is available.we have some significant data issues which need to be highlighted at the outset. Bihar. derived from data sets not intended to address the Hindu/Muslim differentials per se . JAN-DEC 2014 "Differentials across SRCs in MPCE levels in different states are similar to those observed at the national level in both urban and rural areas (see Appendix Tables 8. p. Sachar preferred to concentrate on the proportion of rural SRC households in Monthly Per Capita Expenditure (MPCE) classes. However. "The pattern of distribution of households of SRCs by broad expenditure classes in rural areas (Fig. A large proportion of Muslim. As compared to urban areas.

Sachar’s Appendix Tables 8. we have perforce to assume that these factors do not vary across the two communities. 1-4 HINDU-MUSLIM RURAL HOUSEHOLD COMPARISONS segment of this sample relating to a sub-group cannot be representative of the entire sub-group.2 and 8.3 provide average MPCE levels for each of the States as derived from the NSSO 61st Round data. (Given the above. 2002. reclassified for Sachar in terms of religious groups (SRCs). this gender-age It is generally accepted that in the survey mode weightage (discount) may be expected to have of data collection. These factors . However. these problems can be limited by comparing the ratio between Muslim and Hindu income and expenditure separately for each relevant year and by examining how those ratios have changed over the decade. 87 concentrate on household expenditure as collected by their investigators. In this case. sample of a specific group needs to be drawn from a universe exclusively containing that group. even where the NSSO/NCAER main samples are representative of each State. household. women and older to expenditure in the later one. one would expect the NCAER income data to be underestimates).5 Unless adequate weightage is given to gender and age profile of the households. To be truly representative. the All India estimates for Muslims as given in Sachar are flawed as the Muslim population is concentrated in a handful of States and these All-India Muslim estimates are not statistically representative of the Muslim population. 26 NOS. Moreover.VOL. the Muslim minority is not evenly and proportionally spread across the country.4 members than Hindu households. dis-saving or borrowing and also on private gifts or government assistance. There is also the issue of differences in savings behaviour over the period and between religious groups. the income Second. Thus. Comparison between these two data sets would necessarily underestimate the growth rate estimates of Hindu and Muslim income/expenwhen income of the earlier data set is compared diture and the fact that Muslim households have a higher proportion of children. the NCAER (2001. The Sachar (NSSO) household expenditure data would depend on household income. The NSSO first collected expenditure data for each sample household and then reduced it to per capita figures using household size without weightage for the age profile of the household.the uneven distribution of Muslim population (both interState and intra-State) must be borne in mind while considering the relevant estimates. First. differences between the two religious groups. respondents are likely to significant effect on these weighted estimates of under-estimate or even mis-report their income. data resulting from such extraction of a segment of the sample would not accurately estimate variables of segments of the spatially unevenly distributed underlying population. within each State the Muslim population may not be evenly distributed and may thus be inadequately represented in the extracted Muslim sub-sample. As stated earlier. In the absence of relevant information. while Sachar gives expenditure per Given the differences between per capita capita. Another major issue arises from the need to re-convert the Sachar MPCE data back into household expenditure. These data problems must be borne in mind as the surveys do not allow us to reach firm conclusions on these associated issues. 2003 & /expenditure figures will not represent true per 2004) provide data on annual income per capita estimates. This is the main reason why NSSO surveys One could expect that the greater size of Muslim .

West Bengal. These are further multiplied by 12. Despite the above stated caution regarding the unevenness of the Muslim population distribution within each state. Household size for the two religious groups has been generally falling at different rates and from different base levels. so we have to take recourse to the NSSO [2007a] for this data. It would.88 JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY households would then be offset by these structural factors in the gender-age weighted estimates. out of the 16 States surveyed by NCAER. SC. Kerala. Yet another important issue relates to the social composition of the two religious communities. only nine State reports give figures for Hindu and Muslim income data. Gujarat. The size of the household and its participation in income generating work is a result of household decisions. Karnataka. Christian. 18] to compute the weighted per capita income/expenditure of the two communities. JAN-DEC 2014 For all these reasons.17 (=365/30) to get annual figures. while in 2004 the figures were 961 and 968 respectively [NSSO. Sachar has made comparisons of the General Muslim MPCE with Hindu Scheduled Caste (SC)/ Scheduled Tribe (ST) and Hindu Other Backward Class (OBC) MPCE’s. However. The NSSO estimates for sex ratio (number of females per 1000 males) in rural areas in 1994 was 941 for Hindus and 960 for Muslims. it would be better to deal with household estimates. The general effect of household size and structure on income/expenditure will be indicated in due course. Sikh. well as to the Hindu majority community in varying proportions (Table 1). Further.6 (In the section on "Results". The next issue is that since the NCAER data relates to rural households we can only compare it with the rural sample data in Sachar. Uttar Pradesh. more detailed and micro level analysis is needed to be done using the raw data. is not to contend that household income or expenditure is an exact index of the household’s well being but in economic terms this may be the only available indicator since per capita income/expenditure suffers from the above mentioned weaknesses. seem better to deal with this complexity openly rather than reduce it to this sort of simplistic standardisation. these nine States . p. More critically. unweighted per capita income/expenditure estimates result in favorable intertemporal comparisons if household size has fallen over time. Maharashtra and Rajasthan . Unlike the NCAER reports. . ST and OBC segments among Muslims too. Further. Sachar (surprisingly) does not give the figures for household size of the religious communities. Buddhist and Zoroastrian .the last of these also depend upon differences in socio-economic and cultural influences. however. the differences in household size both between Hindus and Muslims in the same data set have a significant impact on figures expressed in per capita terms. This seems somewhat tenuous given the presence of Upper Caste. These could have a major effect on household income/expenditure together with other factors such as differences in land resources.7 As Sachar itself acknowledged these categories extend to religious minorities Muslim. available employment opportunities and individual capabilities and household preferences . 2007. we have given ratios based on both household as well as per capita figures in Table 3 and discussed the implications). therefore.Bihar. Andhra Pradesh.account (as per 2001 Census) for 80% of All-India Muslim rural population and also 64% of All-India total Muslim population. this nine-state data covers a substantial part of the All India Muslim rural population. As we have noted earlier. other factors are present but those are beyond the scope of this paper and have been dealt with by Sachar. To go back to household expenditure figures we need to multiply the relevant State MPCE figures by the household size data of the NSSO samples to obtain household monthly estimates.

The proper procedure would be to compare each social group with a similar one in a different religious group.8 9.4 0.3 46.8 0.0 100. Assuming that the income levels of SCs.VOL.7 30. In the absence of robust sample data disaggregated by caste and truly representative of the underlying caste population.5 42. Distribution of Religious Population by Caste Categories (All India. which have been selected for special state support. 212]. For our nine states the NSSO/Sachar sample size for rural Muslim households ranges from a low of 106 (Rajasthan) to a high of 1505 (West Bengal). p.0 100. as Sachar does.0 100.3 70.6 15. compare the All-Muslim group with the 89 Hindu-SCs/STs. p.5 32. would negatively affect the average income/expenditure levels of the religious groups of which they are a part.4 2. Muslims have twice the proportion of "Other Castes" (presumably Upper Castes) than Hindus. the official nomenclature of this sub-group is "Backward Class". OBC data and the smaller Hindu household size pushing them upwards .0 30. 2006.0 100. Sachar devotes a whole chapter to Muslim OBCs10 which provides data on this group.all relative to the All-Muslim estimates. This then is a select group of castes or sub-castes from the main and larger backward caste group.9 7. 1-4 HINDU-MUSLIM RURAL HOUSEHOLD COMPARISONS Table 1. STs and OBCs are lower than that of Upper Castes. 26 NOS. . 2004-05) Religion SCs STs OBCs Others All (1) (2) (3) (4) (5) (6) 22. the MPCE’s indicate that Muslim OBCs do better than Hindu OBCs in rural areas (Rs 566 compared to Rs 548) while the reverse is the case in urban areas (Rs 689 compared to Rs 901).8 39. The data indicates that the proportions of Scheduled Castes (SCs) and Scheduled Tribes (STs) in Christian and Buddhist minorities are larger than in Hindu majority.5 33.1 26.8 Hindu-OBCs and HinduGeneral (or "Other". p. the proportion of SCs.0 Hindu Muslim Christian Sikhs Buddhists Jains Zorastrians Others Total 2.7 94.0 100. we cannot. Only then would there be a like-to-like comparison.4 8.0 100.4 3.5 8. We must also bear in mind that Sachar defined its category of Backward Castes as "Other Backward Classes" (OBCs) ‘as listed in the comprehensive list of OBCs prepared by the National and State Backward Classes Commissions and adopted by the Central and State Governments for reservation for various purposes.0 13. The data also indicates that Hindus have a much larger proportion of SCs and STs than do Muslims.7 89. ST. Sachar does not provide caste level disaggregation for Muslims and anyway the resultant sub-sample size may not generate valid data. 2006. All this by itself will affect to varying degrees the differences between the averages by pulling the All-Hindu estimates downwards due to the Hindu-SC.7 6. while they both have similar proportion of OBCs.9 However. 7]. 3]. Similarly.assuming that the upper castes generate greater income/expenditure than the others. while in the Sikh minority they are the same as in the Hindu majority.0 100.’ [Sachar. While the sociological category is that of caste. Sachar gives the All-India MPCE averages for Muslim OBCs and Hindu OBCs [Sachar.0 100. the twice as high proportion of Muslim Upper Castes in the All-Muslim group compared to the Hindu would push the Muslim average upwards .7 Source: Sachar [2006. This would also indicate that the difference between the two OBCs of the two religious groups is mainly in the urban rather than in the rural sector. STs and OBCs.8 100. Despite these issues.1 2.8 22.6 19. They may be either Hindu or Muslim in religious terms.1 0. Upper Caste) groups alternatively.2 24.5 9.9 82.2 41.2 0.0 59. But Sachar has added the Muslim SC/ST groups to the Muslim BC group11 thus biasing the figures downwards and further confusing the interpretation of the data when compared to Hindu OBCs.

13 To reiterate. while Hindu household size there fell in 2004 . Thus any straight comparison across the decade will underestimate the change as noted earlier. STs and OBCs across the country. But given the caste composition of the religious groups cited above this will tend to bring the All-Hindu and AllMuslim household averages closer to each other due to the higher proportion of SCs and STs in the Hindu sample.17 to get annual household expenditure estimates. Rural Household Size STATES (1) Bihar West Bengal Kerala Uttar Pradesh Karnataka Andhra Pradesh Gujarat Maharashtra Rajasthan HINDU MUSLIM 1994 NCAER 2004 61st R 1994 NCAER 2004 61st R (2) (3) (4) (5) 6.2 6. The differences in household sizes between the two religious communities and the changes in household size over time would have had a significant effect on any comparison in per capita terms.3 6. the Sachar/NSSO data does not take into account the uneven distribution of the population of SCs.3 NCAER (various). bearing in mind the data issues referred to earlier.1 5. further.0 6. this paper.7 4. Using All-Hindu category for comparison of income/expenditure with the All-Muslim category will thus be influenced by the main internal differences of the two religious groups which include different caste (and class?) sub groups.2 5. that the NCAER income data is not disaggregated by social categories we will use the Sachar expenditure data relevant to the AllHindu category (including all Hindu social groups) for comparison with the All-Muslim category (including all Muslim social groups) to ensure religious comparability at least. household size is the same JAN-DEC 2014 for Hindus and Muslims in both periods.8 5. In the case of Bihar. As noted earlier. we may not have accurate estimates which can be used for drawing firm conclusions.0 4. However. pp. This permits comparison with NCAER rural household annual income averages. Thus.6 6. Table 2. the difference between Muslim and Hindu household sizes generally varies but Muslim households are larger. These three States apart. NSSO [2007a.3 5.creating a gap. ST and OBC aspects of the data as it will only compound the data problems we already have.3 5. Muslim household size remains constant in Uttar Pradesh and Rajasthan. The resultant averages will be biased downwards depending on the proportions of the sub groups with lower income/expenditure in the religious group. using household figures ensured that we took into account such demographic differences in the groups and their changes over time.12 Thus.2 6.1 4.8 5.1 6.4 6. . when considering the Sachar data relating to these caste groups in isolation.7 5.3 5. But it must be emphasised again that Sachar estimates are for expenditure while NCAER provided income data. These show declines across most states over the decade. we have re-converted the Sachar monthly per capita data to annual rural household estimates by multiplying them with the relevant household size and then by 12.0 5.6 4.3 5.3 6.5 5.5 6. concentrates only on the changes in the ratios between the two religious groups in. The identification of the exact effect requires use of the raw data and application of more sophisticated tools. and between. therefore.2 6.0 4.4 4. For these reasons we will not deal with the SC.9 6. 36-37] Rural household income/expenditure per annum for Hindu and Muslim households in 1994 and 2004 is given in Table 3.4 5. RESULTS Table 2 gives the household size for each community in the nine states.90 JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY Further.7 4.7 4. Given.4 6.6 6.5 5.1 6. each period.

The two alternate ratios for Bihar show no change over the decade as the household sizes have not changed. This also indicates a general and substantial narrowing of the Muslim-Hindu 1994 differentials in the nine states except in the case of West Bengal where there has been a widening of the gap. than that for Hindus. However. The lowest Muslim-Hindu ratio is for Muslim rural households gathered by the 50th Andhra Pradesh at 63% while Gujarat at 71% Round (1994) and 61st Round (2004) of the provides the second lowest. This last aspect is beyond the scope of this paper but it is well to bear in mind the internal dynamics of the household may have significant effect. the analysis in the present paper has been carried out on the basis of household data. though these may be difficult to estimate. This data suggests that Muslim households over the decade have done better than Hindu households especially in Gujarat and Andhra Pradesh (which in 1994 had the lowest ratios) and in Maharashtra. employment and unemployment data. In the 2004 Sachar data. EXPLANATORY FACTORS General rural economic conditions in each state over the decade would necessarily have had an impact on income/expenditure growth both for Hindus and Muslims in that state. It also determines the allocation of resources including consumption expenditure among its members. That analysis is beyond the scope of this paper and must await further work. the average incomes of Muslims are lower than that Table 4 gives the age profiles of Hindu and of Hindus. however.) HINDUS MUSLIMS SACHAR 2004 HH Expenditure (Rs. there are some basic elements .such as household age profile. as the household is the key organisational and decision-making entity. This reflects changes in household size over time and between communities. The remainder have NSSO.) HINDUS % MUSLIM/HINDU Household % MUSLIM/ HINDU Per capita MUSLIMS 1994 2004 1994 2004 (1) (2) (3) (4) (5) (6) (7) (8) (9) Bhar West Bengal Kerala Uttar Pradesh Karnataka Andhra Pradesh Gujarat Maharashtra Rajasthan 22812 18441 26344 27079 27801 25529 30056 30947 26803 21369 17401 29991 24298 23661 16142 21213 26179 34228 29436 34886 48392 39716 30081 29398 39181 35219 40408 27991 31700 62427 39019 33014 33401 44705 38548 46839 94 94 114 90 85 63 71 85 128 95 91 129 98 110 114 114 109 116 94 85 87 90 76 51 65 76 128 95 82 100 94 101 101 104 96 103 In order to show the degree to which changes in ratios are affected by changes from household estimates to per capita estimates we have added two columns on the right of the Table 3. The direction of the change in the alternate ratios is not altered but the magnitudes are.VOL. 1-4 HINDU-MUSLIM RURAL HOUSEHOLD COMPARISONS 91 Table 3. it seems there is no fixed pattern across the households have a higher proportion of workingnine states in 1994. which gives All-India data on Hindu and ratios between 85% and 94% On the basis of this Muslim households. 26 NOS. These indicate that Hindu data. which are The 1994 income data for Kerala and Rajasexamined below based on NSSO data for the than show Muslim average income to be higher entire country. . the average Muslim household expenditure is nearly equal (90%-98%) or greater than the Hindu equivalent in each of the nine states. However. age members than Muslim households. in other seven states. Rural Household Annual Income/Expenditure STATES NCAER 1994 HH Income (Rs.

NSSO (2007a: Table 5. It also suggests the need for higher Muslim household expenditure to support the larger proportion of non-working proportion of their households. Uttar Pradesh. NSSO (2007a: 50-51). Data regrouped. All India Rural Household Age Profile FEMALES 50th Round (1994) Years (1) Below 15 15-54 55 + Total Hindus MALES 61st Round (2004) Muslims Hindus Muslims 50th Round (1994) Hindus 61st Round (2004) Muslims Hindus Muslims (2) (3) (4) (5) (6) (7) (8) (9) 355 542 103 1000 427 494 79 1000 341 560 99 1000 394 530 76 1000 375 527 98 1000 451 470 79 1000 359 539 102 1000 428 493 79 1000 Sources: NSSO (1998: Table 7. Andhra Pradesh and Maharashtra (Table 5). Table 5. Hindus Muslims 61st Round (2004) Hindus Muslims . over the decade there has been an improvement in the proportion of Muslim males employed in five states especially in Gujarat. the proportions of working members have increased in Hindu households and to a lesser extent also in Muslim households. However. A-8). No common pattern As far as the proportion of male population is visible in the changes in the proportion employed is concerned. However. however. There have been. declines in Bihar. A-11.92 JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY JAN-DEC 2014 Table 4. were generally significantly lower than Hindu ones across the nine states and in both periods. A-7. over the decade (1994-2004). the Muslim proportions employed. This will enhance the supply of family labour and income earning potential of Hindu households which have larger proportions of persons in the 15-55 year age-group. A-12). Proportion of Rural Persons Employed (Principal and Subsidiary Status) Per 1000 persons FEMALES 50th Round (1994) Years (1) Bihar West Bengal Kerala Uttar Pradesh Karnataka Andhra Pradesh Gujarat Maharashtra Rajasthan Hindus Muslims (2) 178 212 265 230 442 529 401 484 467 MALES 61st Round (2004) 50th Round (1994) Hindus Muslims (3) (4) (5) (6) (7) (8) (9) 99 104 134 141 272 359 254 317 293 150 194 302 249 470 490 436 480 411 74 131 124 190 348 397 322 280 236 514 571 565 527 607 634 575 555 543 484 514 431 481 559 585 527 498 450 487 591 596 504 635 609 590 572 509 413 541 448 450 600 538 619 476 481 Sources: NSSO (1998: Statement 10). where Muslim proportion exceeded Hindu proportion in 2004.

that is. Table 6 presents the figures (for 2007-08) for Hindu and Muslim household income per annum and Muslim incomes as a percentage of Hindu incomes in the 42 MCD’s. commissioned by ICSSR and titled "Overview of the Findings". Buddhists and Zorastrians (Parsis) as notified under Section 2 (c) of the National Commission for Minorities Act. the data is 4 years later than that used in the Sachar Report. Islamia. New Delhi. or higher. But despite higher household size. fewer proportion of both male and female members gainfully employed. In the other four states (Bihar. This indicates an uneven pattern from which no firm conclusions can be drawn about the condition of the Muslim community across these 42 districts. 26 NOS. Under the overall coordination of the ICSSR. Guwahati. In another 6 districts Muslim incomes are 60% to 70% of Hindu incomes. DISTRICT LEVEL DATA As a policy reaction to the Sachar Report. . the Government of India undertook to accelerate development in "Minority Concentration Districts" (MCDs) among other steps. A report. The point is illustrated by the fact that of the 89 MCD surveys only 42 present average annual household income for Hindus and Muslims. 1-4 HINDU-MUSLIM RURAL HOUSEHOLD COMPARISONS 93 The proportion of Muslim women employed is very much lower than that of Hindu women across the nine states and in both periods. These surveys were conducted by independent organisations using a common methodology prescribed by the Ministry. Sikhs. Kerala.16 The surveys commenced in 2007 and were completed in 2008. Uttar Pradesh.18 deals with some of the issues and shortcomings. Yet household income data is critical by itself and can serve as a baseline estimate for measuring the effectiveness of the investment in the MCDs. 1992.15 However. reputed research institutions in the states. (v) the employment and consequent poorer income Aligarh Muslim University and (vi) Jamia Millia generation. In 16 MCDs. Karnataka. smaller proportion of members in working ages.17 This is a vital omission.VOL. their quality is somewhat variable and not all are comparable with each other. and (iv) Institute for changing pattern for low Muslim female Human Development (IHD). however.let alone 25%). This is indeed creditable and encouraging. the survey work was conducted by (i) Centre for Studies in Social Sciences (CSSS). The MCD surveys were guided by the Indian Council of Social Science Research (ICSSR) and the work was entrusted to Despite this. Neither the MCD reports nor the "Overview" deals with income differential data even when collected. Muslim household incomes are equal to. Thus. The lowest three are for Sirsa in Haryana (51%). Uddham Singh Nagar in Uttaranchal (52%) and Mamit in Mizoram (58%). the 89 MCDs which were selected by the Ministry of Minority Affairs are mostly those districts where Muslims have a large share of the population (though as can be seen there are many districts where their share is well below the national average . in 9 districts they are between 70% and 89% of Hindu incomes and in 8 districts Muslim incomes are 90-97% of Hindu incomes.14 The MCDs were to be selected on the basis that they had substantial proportion (at least 25%) of religious minorities. Lucknow. Maharashtra and Rajasthan). Christians. Andhra Pradesh and Gujarat. than Hindu ones. there has been an increase in the proportion of Muslim women employed over the decade in five states .West Bengal. Muslims. Muslim rural households on average have caught up in respect of employment with Hindu rural households over the decade. The Ministry also commissioned base-line surveys of MCDs preparatory to funding development of these districts. (ii) Giri Institute for Development Studies (GIDS). Kolkata. the proportion has declined. But. These vary considerably. as Table 5 also indicates. (iii) OKD Institute for Social Change This seems to be a firm though gradually and Development.

P. 2008.) . pp: 41 & 52 http://www. Bijnor UP Prasad. 2008a.pdf 21.ics sr.1 52985 35482 67 10. Baduan UP Fahimuddin. pp:22-40 & 50 http://www.3 23737 26341 111 7. Lucknow UP Kumar. 2008a. pp: 20-33 & 53 http://www.pp::25-44 & 55 http://www.icssr.7 30948 25994 84 6. 2008.ic ssr.pp: 24 & 64.ics sr. Muzzafarnagar UP Prasad.icssr.pdf 24.pdf D%20M%20Diwakar.94 JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY JAN-DEC 2014 Table 6.1 61017 66457 109 9.5 34068 45557 134 11.%20Nagar %20-%20BK%20Bajpai.i cssr.pdf 21.pdf 22.pdf 34. Bahraich UP Garia. http://www.ic ssr. Shrawasti Nayak.1 30970 24730 80 15.pdf. 2008a. Barabanki UP RC%20Tyagi. Bagpat UP Tyagi.1 43665 37975 87 5. 2008. JP Nagar UP Bajpai.i cssr. pp:18-31 & 46 http://www .org/District%20Kheri%20-%20A%20Jo shi. Average Annual Household Income in MCDs 2007-08 (Rupees) No. 2008.pdf.0 32195 26306 82 4.pdf 39. 2008. Lashmipur Keri UP Joshi.7 46333 41341 89 3. pp: 33 & 52-53 http://www.pdf 20. pp:18-34 & 64 http://www.i 0Garia. Bulandshar UP Singh.pdf 38. pp: 27-43 & 55 Lucknow.pdf t%20(Baseline).org/District%20Badaun%20-%20F ahimuddin.icssr . icssr. icssr.pp:19-35 & 66 Baseline).pdf %5B1%5D. A%20Jafri.icssr . 2008.S. 2008. District State Source19 % Muslim Population Hindu Muslim % M/H (1) (2) (3) (4) (5) (6) (7) (8) 1.4 31712 39406 124 13. 2008a. 2008.pp:18-37 & 51 http://www. 2008.pdf 21.pp:18-37 & 51 http://www.ics 20Garia. pp:14 & 36 http://ww 20Nayak.6 41876 37107 89 UP (Contd.4 50987 50842 100 14.i cssr. Saharanpur UP Tyagi.7 44735 30498 68 Balrampur UP Diwakar.%20Tiwari.5 40625 41637 102 12. Moradabad UP Tiwari. pp: 18-33 & 47 http://www. 36.pdf 19. Siddharthanagar UP Garia.8 30153 28224 94 2. 2008a.pp: 29 & 48 http://www. 2008.

6 52013 32002 62 32.pp:12 & 26 http://www. 2008a.pdf 49.5 55461 48445 87 34. 2008. pdf.icssr.7 42840 37636 88 19.pp::36 & 56 http://www. pp:16 & 21 www.7 25148 33398 133 22. Gumla Jharkhand Sharma & 5D.pp:14 & 26 http://w ww.pp: 14 & 17 ht tp://www.pdf 0Fahimuddin. U Singh Nagar Uttarkhand Sudan.pp: 15 & 27 http://www. pp:16 & 26.icssr. 2008. Araria Bihar Pankaj.pdf 10. Washim Maharashtra 6. http://www. 2008. Hardwar Uttarkhand Mamgain.4 32804 29951 91 16.pp: 15 & 31 http ://www.icssr.icssr.9 27534 25278 92 .) No.pdf 32.icssr. Tiwari%5B1%5D. 2008. 2008.4 33051 35216 107 31.pdf 20final%5B1%5D.pp: 13 & 21 www. &18 Parbhani Maharashtra Singh & Sudan.pdf 21.pdf 9.pdf 44.1 62503 37864 61 33.icssr.pp:16 & 28 www.3 33174 32149 97 21.icssr.pdf Bhaskaran.VOL.icssr .org/Haridwar_Final. 2008.pdf Accessed 15th July 2011 32.pdf 33. District (1) State (2) (3) Source19 % Muslim Population Hindu Muslim % M/H (4) (5) (6) (7) (8) 16.6 18996 29004 153 28. Bareilly UP Fahimuddin.pp: 18 & 27 http:// www. 2008a.8 26065 27966 107 33565 33426 100 30. http://w ww.7 29070 38438 132 29.pdf kur%20final%5B1%5D.1 32055 28891 90 26.pp:15-25 & 43 http://w ww. Kishangunj Bihar Sudan.pdf 37.pdf Bhaskaran.icss r.1 50898 32876 65 (Contd. Hingoli Maharashtra 6. 2008. P-Chamaparan Bihar Pankaj & Mishra. Katihar Bihar Pankaj.pdf 6.4 32065 31012 97 25. 1-4 HINDU-MUSLIM RURAL HOUSEHOLD COMPARISONS 95 Table 6.pdf 21.1 34291 38020 111 pp: 16 & 27 http ://www. Bhuldana Maharashtra Bhaskaran. pp: 15 & 27 http:// www. Purnia Bihar Pankaj & Poornima 2008. 26 NOS. ( 2008a. Rampur UP Tiwari. pp: 13 & 25 http://www.icssr. 2008. 2008b.icssr. Ranchi Jharkhand Dayal & Singh.icssr.pp:17 & 30 www. pp: 14 & 25 http://w ww. Dharbhanga Bihar Pankaj & l%5B1%5D. 2008. 2008.1 42940 45699 106 2008. Sitamarhi Bihar Mishra &Singh.icssr. Sahibganj Jharkhand Sudan & Bhaskaran. 22.icssr.icssr.2 84796 43985 52 2008a.pp: 13 & 26. Pakur Jharkhand Dayal.

org/Waynad%5B1%5D. this exercise brought out the difficulties in using available data and in making definitive statements about the trend and the variation between incomes/expenditures ratios of Hindus and Muslim households.pdf negl 86375 44063 51 74.9 83079 76527 92 12. The Government of India with its enormous resources and control of the NSSO should initiate serious. pp: 16 & 28 www. (Contd. In the process.ic ssr. overtaking rural households of the Hindu majority. 2008c. Bidar Karnataka 39.pdf Upendranadh.2 48324 43539 90 The interpretation of this data itself poses problems and the analysis is tentative. This subject is too serious to be treated in such an ad hoc manner. 2008.3 51203 35254 69 26.pp:23 & 43 http://w ww. 2008. which led to the mistaken and disoriented policy decision to concentrate investment on rural areas rather than on urban areas where the Muslim disadvantages are significant (as indicated in Sachar itself).org/Gulbarga%20final%5B1 2008.icssr.pp: 23 & 49 http://ww w.pp: 20 & 36 http://www. Mewat Haryana 37. rural Muslim households seem to be catching up and. pp: 20 & 26 http:// www. The NSSO with its international reputation and experience at data collection should be entrusted with the job of conducting regular specifically designed surveys with the aim of . it is difficult to conclude that Hindu-Muslim rural differentials in income/expenditure have a fixed pattern across States and over time.icssr. In order to do so. It seems clear that it is not enough to extract data in the manner in which the Sachar Report has done and to draw conclusions. 2008. Wynad Kerala Source19 % Muslim Population Hindu Muslim % M/H (4) (5) (6) (7) (8) Mehta et al. Lawngtlai Mizoram 42.pdf More robust surveys are required to enable clearer understanding of the dynamics of household income/expenditure decisions of Hindus and Muslims based on their preferences and available resources. District (1) State (2) (3) 35. Sachar and MCD survey sources.pp:16 & 26 www. Our primary purpose was to examine the trend of income /expenditure ratios for the Hindus and Muslims in 1994 and 2004.icssr.p df Deogankar.6 49150 40251 82 1. perforce. Leh J&K No. Gulbarga Karnataka Prasad. 2008b.icssr.pp : 14 & 25 www.pdf Sudan. we had to deal. consistent and comparable studies over time of the progress (or lack of it) of the Muslim minority community with reference to the Hindu majority.2 34798 48580 140 12. with these different data sets. 2008.pp: 13 & 26 www. Sirsa Haryana 36.pdf Sudan.96 JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY JAN-DEC 2014 Table 6. 2008a.icssr. Mamit Mizoram Sudan & Mamgain.5 44446 56802 128 SUMMING UP After examining the above data from the NCAER. If anything.icssr. in many cases. Even the selection of MCDs seems to have been misplaced and misdirected as indicated by the initial survey data showing high levels of Muslim household incomes relative to Hindu households in nearly 24 of the 42 MCDs.8 58006 33627 58 0.

p. "The baseline surveys in the identified MCDs were to bring out the following: (i) A gap analysis of availability of infrastructure like schools. 9 Incidentally. which in turn could lead to well-directed remedial measures. Table 5. seems c. This data will then enable both policy-makers and civil society to assess the direction and pace of the progress of minority religious communities in different states and districts within them.indianexp ress. Pp. Combining SC and ST data into one groups when the two populations are significantly different in economic status is unwarranted. 5. Ahmad [1967. See endnote 4 moma/files/guideline.VOL. Comparison of expenditure and income figures for 2004 from two different data sets indicates that income estimates arehigher than expenditure estimates by about 54% for Hindus and 40% for Muslims [Shukla. 1-4 HINDU-MUSLIM RURAL HOUSEHOLD COMPARISONS examining the progress of the Muslim minority taking into account the spatial distribution of its population. Also see Aggarwal [1966. This also prevents meaningful inter farm comparisons. It was also a period of liberalisation and transition to a market economy. Good intentions alone will not deliver results. 14.09. p. it indicates a relatively faster Muslim progress and consequently a gradual closing of the initial gap over the decade 1994-2004. As every NSSO Report states: "The size of a household is the total number of persons in the household.10. p. 114] for Maharashtra Muslim moma/files/ ks-off-war-of-words/809657/0) (Accessed 7. But this paper also concludes that more serious and nuanced efforts need to be made by NSSO to get reliable data to facilitate correct interpretation.even my modest collection contains over 150 entries. pp. 10. This paper does not substantiate the currently held notion of fixed and firm Muslim income/expenditure differences compared to the Hindu majority in rural areas. Sachar [ (Accessed 21. . These surveys need to be robust.2014). 189-216]. p. NOTES 1. 38]. 1976]. nuanced. See http://minorityaffairs." 6. 11. A similar problem emerges in the comparison of farm income. Pp. which has disrupted many patterns of traditional rural livelihoods. is quoted as saying that the ‘Sachar Report is not the Koran which cannot be questioned!’ (http://www. 16. Sachar’s rural estimates for All-Muslims in Delhi is based on a sample of only two householdsillustrating the need for adequate degrees of freedom for credible averages and also the need for statistical significance tests. 47) for Bihar Muslim data and Shaban [2011. Most analysts tend to use econometric tool such as regressions. With different extent of land with differences in size and nature of its cultivated and irrigated portions. See Asian Development Research Institute (ADRI) (no date given. principal component or factor analysis to separate out the varying effects of these factors on farm income [Pingle. Table 1.3.2014). (See interesting Cartograms: http://realitycheck. 12. 2002-2003.12. 159-162] and Basant [2007]. The academic and other studies on the condition of Muslims are extensive .9. by not taking into account land endowments and farmer’s crop decisions. They should have methodological consistency so as to be comparable over time.2014). "The cross-sectional pattern with respect to consumption and poverty differentials are analysed at using the most recent NSSO 61st Round data with a reference period of July 2004 to June 2005" [Sachar 2006. health centres. 887-891] deals with the categories of Ashraf and Ajlaf in the Muslim community. p.2014). any attempt to reduce farm income to per acre terms causes loss of detail. 38-40]. See http://minorityaffairs. Also see Shukla [2010. Table 2.pdf (Accessed 30. On the other hand. 26 NOS.22. 15. (iii) Identification of income generating activities in which the villagers have comparative advantage. 7. 8. then Union Minister for Minority Affairs. disaggregated and location specific and with adequate sample sizes to enable statistically significant comparisons. 151]. ICDS centres and drinking water supply. 2. Ibid (p. That this should happen in the decade under review is especially heartening for it was a period of low and erratic growth of agricultural sector. This is a result of the efforts of individual Muslim households and needs to be recognised and lauded. 2010. 97 3.wor dpress. Salman Khursheed. robust analysis and sound understanding of the economic status of the Muslim minority in relation to the Hindu majority.9.pdf (Accessed 30. (ii) A gap analysis of housing and sanitary toilets. 13. cropping intensities and cropping pattern. 5).

National Sample Survey Office (NSSO). accessed on 21. 2011.D. UK. New Delhi.1. 512. New Delhi. Socio-Economic and Educational Status of Muslims in Maharashtra.icssr. For all 89 reports see : http://www.. 2003. National Sample Survey Office (NSSO).org/minority-dist. New Delhi.htm). p. Thesis. Glasgow University. Report No. Economic and Political Weekly. Perceived Adequacy of Food Consumption in Indian Households 2004-2005. Patna.htm (Accessed 7. 2001. Unpublished Ph.141. banking facility.htm Accessed on 11th May 2011. 521. All accessed on 15th July 2011. West and Central India: Human Development Report. ARDI. p.09. 2004. Oxford University Press. New Delhi. Caste in a Different Mould.Problems of Cultural Integration".1. Some Methodological Aspects of Cost-Benefit Analysis of Irrigation Projects . North India: Human Development Report. etc. Seminar. National Council of Applied Economic Research (NCAER). Indian Council of Social Science Research.. 2007.56. Sunil Jain and Preeti Kakkar. Government of India. p. Akalank Publications. September 10. REFERENCES Aggarwal. Table A 11. Mumbai. Shaban. New Delhi. 2003. 31 January. Government of India. Table A2.2014). Ministry of Statistics & Programme Implementation. March. Rajesh. Table A4.2014. Ministry of Statistics & Programme Implementation. High Level Committee to Report on the Social. 1976. Rakesh. Government of India.A State Report. June. 2010. Tata Institute of Social Sciences (TISS).52. Economic and Political Weekly. National Council of Applied Economic Research (NCAER). p. p.icssr.2005 Report No. Partap C. Shukla.icssr. Ministry of Statistics & Programme Implementation.1. http://www. A. 1967. BS Book. Table A2. NSS Report No. 61th Round. Table A10. "The Ashraf and Ajlaf Categories in Indo-Muslim Society". 2010. 272. New Delhi For all 89 MCD Baseline Surveys see http://www. 373.A Study Commissioned by the Government of Bihar. Ahmad. no date given.1. 19. East India: Human Development Report. "A Muslim Sub-Caste of North India .c om/2007/569/569_rakesh_basant.10. markets. Economic and Educational Status of the Muslim Community of India (Sachar). Oxford University Press. . 2002. May 13. Pingle. http://www.1. Oxford University Press.. 2007. "Diversity among Indian Muslims".Understanding the Discrimination. ITls. South India: Human Development Report.1.htm 18. Gautam. 1993-94.1. 1998. 2002. Asian Development Research Institute (ADRI). 50th Round.A Case Study of the Telangana Region of India. National Sample Survey Office (NSSO). New Delhi. Socio-Economic and Educational Status of Muslims in Bihar. 2006. 2007a. 2001. National Council of Applied Economic Research (NCAER). Government of India. Oxford University Press. February.india-seminar. survey conducted in 2001-2002). Employment and Unemployment Situation in India. 2007. Basant. Table A7. Imtiaz. Employment and Unemployment Situation in India. JAN-DEC 2014 National Council of Applied Economic Research (NCAER). which will provide the missing links that can act as catalyst" (http://www.98 (iv) JOURNAL OF INDIAN SCHOOL OF POLITICAL ECONOMY A gap analysis of critical linkages like rural roads. Table 10.pdf. 2004 . 61th Round. 438.