August 09, 2015
Array

SECC Excludes More Than It Includes Amidst High Rural Distress

Smita Gupta

COMMISSIONED by the UPA government in 2011, the Socio Economic Caste Census (SECC) was to collect data on caste, urban and rural socio-economic status. The preceding BPL census was undertaken in 2002, which collected information on 13 indicators for every rural household. Households were selected on the basis of their total marks, the cut-off being the total marks at which the number of BPL households in a state equalled its poverty estimate as determined by the Planning Commission. This process was problematic, procedurally and conceptually. It suffered from arbitrariness and was full of errors of wrong inclusion and exclusion.

The impression created at the time was that objective, fair and transparent criteria would be used to determine the list of those excluded and included, and the SECC was to become the comprehensive database for this purpose. The clearly stated policy aim was to improve the identification of beneficiaries of welfare schemes. 

 

DELAYED, PARTIAL

AND UNVERIFIED DATA

Unfortunately, the SECC data has belied this expectation. Much delayed, only the incomplete rural data has been released, partially. Despite the use of state-of-the-art technology, the urban data is still pending, and the government seems spooked by the caste data apparently because the findings are politically contentious. The government cited the prevalence of huge errors in the caste census to explain this delay. It claimed that the data has over 8 crore mistakes, of which 1.46 crores were yet to be rectified. This is a cause of grave concern. Why and how did so many errors creep into this state-of-the-art survey?

Out of the 640 districts covered, data for 628 is in a draft form, and only for 277 is it final. Moreover, 50 percent of the major states have inconsistent and incomplete enumeration with data missing for a few districts each. Of the 35 states and union territories, the ministry of rural development has not published final lists for 21.

Since it is not governed by the Census Act, 1948, the strict requirements of confidentiality do not apply to the SECC. The data was supposed to be subject to verification by the households themselves, as also by gram panchayats and gram sabhas after putting up the data in public places. This process is essential to finalise the list, but as yet it is incomplete and unclear how long it will take.

 

DUBIOUS IDENTIFICATION

METHODOLOGY

The SECC identifies priority households on the basis of a three-pronged exercise. The first is compulsory exclusion of households based on fourteen criteria. The second is an automatic inclusion of households based on five criteria; and there is a gradation of deprivation of the remaining households on the basis of seven criteria. Regrettably while the criteria to exclude are extremely liberal and extensive, perversely the criteria to include are too narrow and restricted. This has resulted in compulsory exclusion of about 40 percent or 70.5 million and automatic inclusion of only 0.96 percent or 1.65 million rural households. Of the total 179.1 million rural households, this leaves 106.9 million households who are to be ranked and graded on the basis of the extent of deprivation. There is a strange discrepancy in the data because there are 20 million rural households who have neither been automatically excluded nor report any of the deprivations listed.

a.                    Deprivation

% of Deprived Households with deprivation criteria (ALL CASTES)

1

Only one room with kucha walls and kucha roof

13.25%

2

No adult member between age 16 to 59

3.64%

3

Female headed households with no adult male member between age 16 to 59

3.85%

4

Disabled member and no able bodied adult member

0.40%

5

SC/ST households

21.53%

6

No literate adult above 25 years

23.52%

7

Landless households deriving major part of their income from manual casual labour

29.97%

 

 

The socio-economic status of a household is calculated on the basis of the seven indicators of deprivation given in the table above. The SECC data will be used to score and “rank” a household on a scale of 0 to 7 on the basis of the number of “deprivations” it has from the seven.

This methodology is frankly absurd. For example, even though adivasi households are the most disadvantaged section of rural India, many possess (may not hold a title) a little bit of land, dry and often unproductive, and a mud house with at least two rooms, and are unlikely not to have an able-bodied male adult aged between 16 and 59 years. As a result they are likely to have a score of one or less and will be excluded.

In fact, some of the deprivation indicators should be parameters for automatic inclusion. Even if 3 of the 7 criteria of deprivation, namely, single room kucha dwelling (3.25 percent of rural households), SC-ST households (21.53 percent of rural households), and landless households primarily dependent on manual labor (30 percent of rural households) are used for automatic inclusion it would pass the test of robustness and fairness.

b.                    Compulsory Exclusion

% OF EXCLUDED HOUSEHOLDS (ALL CASTES)

1

owning motorised two/three/four wheelers/fishing boats

20.69%

2

owning mechanised three/four wheeler agricultural equipments

4.12%

3

kisan credit card with the credit limit of Rs 50,000 and above

3.62%

4

any member as government employee

5.00%

5

non-agricultural enterprises registered with government

2.74%

6

 any member earning more than Rs.10,000 per month

8.29%

7

paying income tax

4.58%

8

paying professional tax

4.58%

9

three or more rooms with pucca walls and pucca roof

18.52%

10

owning refrigerator

11.04%

11

owning landline phones

3.72%

12

owning 2.5 acres or more irrigated land with at least one irrigation equipment

4.27%

13

owning 5 acres or more land irrigated for two or more crop seasons

3.02%

14

owning 7.5 acres or more land with at least one irrigation equipment

2.27%

 

Households with at least one exclusion

39.39%

 

The two criteria which have resulted in the greatest exclusion are ownership of a motorised 2/3/4 wheeler or fishing boat, and, pucca house with 3 plus rooms.  The first problem is the clubbing together of all motorised vehicles and boats, considering that 17.43% of all rural households own motorised 2 wheelers. How can ownership of a motorised 2 wheeler be put on the same footing as a motorised 4 wheeler (2.46 percent households)?  The housing indicator is also not very appropriate because in rural India many nuclear families occupy a single dwelling – and therefore having more rooms may not reflect greater prosperity. Furthermore some caste and religious groups, like Muslims, have larger joint households and therefore more rooms (joint families with many nuclear households in a common kitchen or homestead) but are actually still very poor.

c.                    Automatic Inclusion

% OF INCLUDED HOUSEHOLDS

 

1

Households without shelter

0.07%

2

Destitute/ living on alms

0.31%

3

Manual scavengers

0.05%

4

Primitive tribal groups

0.47%

5

Legally released bonded labourers

0.06%

 

HH with any one inclusion criteria

0.96%

 

As far as the inclusion criteria are concerned, it is obviously a cruel joke. How else can we explain the very small coverage despite the concentration of poverty within certain occupational, caste, and social groups? There is gross under-enumeration of the poorest households. Compare the automatically included (the homeless, the destitute, manual scavengers, PTGs and legally released bonded labour) households which total 1.65 million to the 25 million Antyodaya card holders. Census 2011 reports 7.4 million rural homeless households. While Census 2011 reports 66 million households living in one room, SECC has identified 23.7 million. The ministry of labour in 2014 says that 5 percent of all households have no worker aged 15 years and above, which is only 3.8 percent in the SECC. A Lok Sabha report counts 7,70,000 manual scavengers and their dependents, while SECC reports only 1,80,000. The number of manual scavengers is underestimated, even as the 2011 census reports 2,15,885 dry latrines in Assam, Andhra Pradesh, Tamilnadu and Manipur, while the SECC shows only 653 manual scavengers in these states (Mihir Shah). Evidently, the poorest have not been recognised and surveyed by the SECC.

 

MAIN FINDING:

SEVERE RURAL DISTRESS

The problems notwithstanding, available rural SECC data reveals very high levels of deprivation and severe rural distress. 56 percent of rural households are landless (ownership of land less than 0.01 hectare); 51 percent are dependent on unskilled manual labour for income.; 40 percent of those primarily dependent on unskilled manual labour are  small and marginal farmers; a meagre 30 percent depend primarily on agriculture for livelihood; About 50 percent people are either illiterate or educated below the primary level; Under 5 percent own agricultural equipment; In over 90 percent of rural households, the top earning family member makes less than Rs 10,000 per month; the Scheduled Tribes are the most deprived social group and the rain-fed eastern and central parts of India are the most underprivileged regions.

 

DO NOT USE SECC TO TARGET

WELFARE SCHEMES & NFSA

In conclusion, the main inference from this exercise is that the data and methodology for identifying eligible households for welfare schemes is deeply flawed. In this situation welfare schemes and benefits to the poor cannot become dependent on this selection process. This is especially because no matter what nomenclature is used for the identified households (BPL, priority, etc.) it ends up being a binary process, insofar as it allows each household to either be included or completely excluded from all welfare schemes. Even though the SECC data is useful to realise the full extent of deprivation and distress in rural India, it cannot and must not become the basis for targeting government schemes, in particular the National Food Security Act. The government would do well to only exclude income tax payers, and provide benefits to everybody else.