Estimating the Impact of India’s Aadhaar Scheme on LPG Subsidy Expenditure
In 2015, the Indian government introduced nationwide changes to the system of domestic liquefied petroleum gas (LPG) subsidies, replacing integrated consumer price subsidies with bank transfers through the Direct Benefit Transfer for LPG (DBTL) scheme1. The scheme prioritized linkage with the (non-statutory) Aadhaar identity programme, while providing for a non-Aadhaar linked method of registering for the scheme - reflecting the Supreme Court’s order that Aadhaar enrollment could not be made mandatory for the provision of public services. On March 3 2016, the government tabled legislation in the lower house of parliament (the Lok Sabha) to provide statutory backing to the Aadhaar programme. Framed with the stated objective of improving subsidy targeting and delivery, the legislation has been introduced as a money bill—removing the ability of the upper house (the Rajya Sabha) to exercise full parliamentary oversight.
Within the complex and polarized debate about the relative merits and risks of Aadhaar, the question of the net effect of the program on subsidy expenditure is a key area of discussion. In an article on March 9th 2016 advocating for the rapid passage of the legislation, ex-Chairperson of the Unique Identification Agency of India (UIDAI), Nandan Nilekani, stated that “we have…visible evidence of [Aadhaar’s] utility with Rs 10,000 crore-plus [USD 1.5 billion] savings in LPG subsidy in one year alone”. This followed an earlier statement in December 2015, in which Nilekani reportedly stated that “the government has…saved USD 2.5 billion in one year by linking LPG users to Aadhaar numbers”. On March 11 2016, Finance Minister Arun Jaitley was reported as stating in the Lok Sabha that “targeted subsidy through Aadhar cards of LPG consumers had resulted in savings of over Rs 15,000 crore [USD 2.2 billion]”2.
Are these figures correct? Did the integration of Aadhaar within the DBTL programme result in significant fiscal savings in the current financial year? IISD has previously analyzed the actual fiscal effect of introducing DBTL in FY 2014/15, demonstrating that even without fully accounting for the costs of implementation to government or beneficiaries, the introduction of the scheme generated a net fiscal cost within the financial year. In addition, we have analyzed the specific additionality of the Aadhaar scheme within the re-launched DBTL programme, showing that the maximum gross saving in FY 2014/15 from integrating Aadhaar (and before accounting for any costs) was approximately Rs.12-14 crore (USD 2 million).
Using publicly available information, and applying the same methodology as adopted to calculate additionality in FY 2014/15, it is possible to provide an estimate of the fiscal impact of integrating Aadhaar within the DBTL programme in FY 2015/163. This is calculated by applying the recorded deduplication rate achieved through Aadhaar integration4 to the total number of Aadhaar-seeded LPG connections5, then applying the average monthly consumption rate per active connection6 and the monthly (variable) subsidy rate per cylinder7 (including both direct budgetary subsidy and uncompensated Oil Marketing Company (OMC) costs) to arrive at a total ‘avoided’ subsidy expenditure figure due to Aadhaar (before accounting for any costs of implementation).
Putting aside the broader questions regarding Aadhaar as a public policy tool, our calculations indicate that, instead of resulting in significant savings, the net fiscal impact of integrating Aadhaar into DBTL in the current financial year was significantly negative, and that expectations of substantial net savings in subsidy expenditure from the introduction of the Aadhaar scheme are likely to be misplaced.
This demonstrates that the integration of Aadhaar within the DBTL programme resulted in total savings on LPG subsidy expenditure in FY 2015/16 of approximately Rs. 121 crore (USD 18 million)—less than 1% of the estimated subsidy saving figure of Rs 15,000 crore (USD 2.2 billion). In addition, to obtain an accurate assessment of the net fiscal impact of integrating Aadhaar, this figure should then be offset against the costs of implementation. For example, assuming an average enrolment cost of Rs.120 per person, the marginal cost of enrolling just those customers with Aadhaar-seeded LPG connections as of March 1 2016 (i.e., not including the costs of non-LPG linked enrolment, or any of the costs associated with implementing and operating the DBTL programme itself) was approximately Rs. 1,343 crore (USD 200 million)—ten times greater than the fiscal saving from integration into DBTL in FY 2015/168. An accurate cost-benefit appraisal would also include the direct and indirect costs imposed on beneficiaries.
Putting aside the relative merits of Aadhaar as a public policy tool, our calculations indicate that, instead of resulting in significant savings, the net fiscal impact of integrating Aadhaar into DBTL in the current financial year was likely to be minimal, and that expectations of substantial net savings in subsidy expenditure from the introduction of the Aadhaar scheme may be misplaced.
1. Also known as ‘Pratyaksh Hanstantrit Labh’ (PAHAL).
2. This figure was also repeated on the same day by IT Minister Ravi Shankar Prasad in the Rajya Sabha in support of the bill.
3. It is possible to calculate this figure prior to March 31 2016 as the relevant monthly figures for Aadhaar-based deduplication and subsidy expenditure are determined on the basis of the previous month’s data.
4. Government figures have consistently demonstrated that the maximum number of potential duplicates identified in LPG databases through Aadhaar-based de-duplication is approximately 1% (or less) of total connections assessed. In its July 22 2015 affidavit to the Supreme Court, the Ministry of Finance stated that of 8.08 crore IOCL connections assessed, only 8 lakh potential duplicates were found. In its report on DBTL, the Dhande Committee had previously stated that “[i]n the 291 districts covered under DBTL, 6.18 lakh duplicate connections were identified out of over 40 million [actually 44 million] LPG consumers who provided their Aadhaar numbers”, with an “assum[ed]…50% duplication rate” (a deduplication rate of 0.7%). In the Phase 1 districts of the original DBTL scheme (within which Aadhaar linkage was mandatory), 45,000 customers with two connections were identified out of a total of 5 million assessed, resulting in 22,500 connections being blocked (a deduplication rate of 0.45%) . These figures are additionally corroborated by district-level data collected as part of IISD’s field assessment of the Mysore DBTL pilot programme (details available on request), and may be compared against the deduplication rate of 18-19% achieved through non-Aadhaar based methods (as outlined in IISD's February 2016 policy brief) achieved at a fraction of the equivalent cost.
5. The number of Aadhaar-linked connections was reportedly 8.49 crore as of March 31 2015 and 11.19 crore as of February 29 2016, with a linear rate of increase (and associated identification and blocking of duplicates) assumed.
6. Based on the average rate of subsidized consumption per active connection in FY 2014/15 (7.4 cylinders).
7. As reported by the Ministry of Petroleum and Natural Gas (MoPNG).
8. Note that in a comprehensive cost-benefit analysis this figure would also be offset against savings achieved through the introduction of the initial DBTL scheme in the previous two financial years, and savings (calculated at a maximum of Rs. 12-14 crore) from the reintroduced DBTL/PAHAL scheme in FY 2014/15.