Irrespective of the Petroleum Ministry's inflated claims, during FY2015, actual consumption of subsidised LPG cylinders increased by 18% while the same for unsubsidised refill fell 19%. Even the government's so-called saving claim contains simple arithmetical error of Rs1,205 crore
The Central Government, especially the Ministry of Petroleum & Natural Gas (MoPNG), while patting itself for saving thousands of crores due to its direct benefit transfer (DBTL or PAHAL) scheme, continues to show misleading figures. So much so, the clarification issued by the Ministry too contains arithmetical mistake. Not only this, according to the data available in public domain, contrary to the Ministry's claims, actual consumption of subsidised LPG cylinders has gone up by 18% during FY2014-15.
In a release on 12 October 2015, the MoPNG had said, "If we take into account the quota of 12 cylinders per consumer and the average LPG subsidy of Rs336 per cylinder for the year 2014-15, estimated savings in LPG subsidy due to the blocking of 3.34 crore accounts work out to Rs14,672 crore, during that year."
So let’s do the calculations...
12 (refill) x 3.34 crore (blocked connections) x Rs336 (average subsidy for each refill)= Rs13,467 crore
This means, even revised figure of the estimated saving by the government is inflated by Rs1,205 crore (Rs14,672 crore - Rs13,467 crore = Rs1,205 crore).
As per the data from the MoPNG, during FY2014-15, total subsidised consumption increased 18% or 16.6 crore cylinders. At the same time consumption of unsubsidised cylinders fell by 19% or by 4.3 crore cylinders. In FY 2014-15, total LPG consumption grew by over 10% relative to FY 2013-14 - a sharp increase on the previous year's growth rate of 4.7%. It must be noted that despite sharp increase in consumption of subdised LPG cylinders, overall subsidy expenditure of the government decreased. This happened due to rapidly falling international oil prices from late 2014 onwards, leading to a 65% decrease in recorded LPG unit under-recoveries between November 2014 and February 2015. Data from MoPNG shows per unit under recoveries for LPG fell to Rs143.7 in March 2015 from Rs506.1 in April 2014.
The significant increase in overall LPG consumption in FY 2014-15 was due almost entirely to a sharp rise in the consumption of subsidized household LPG in FY 2014-15, increasing by almost 18% (the equivalent of 166 million 14.2kg cylinders) relative to FY 2013-14 (see Figure below).
Due to the UPA administration's relaxation of the per household allocation, and the current government's subsequent failure to reinstate a realistic per household quota, subsidized consumption then rose sharply in FY 2014-15 - increasing by the equivalent of 166.3 million cylinders relative to FY 2013-14 (as detailed in Figure above) - while unsubsidised consumption fell to 13.9% of total consumption.
In addition, to the significant increase in subsidized consumption, and despite a strong increase in overall LPG consumption, total unsubsidized consumption (including all categories) fell sharply from its previous high in FY 2013-14, declining by over 19% (the equivalent of 43 million 14.2kg cylinders) (see Figure below).
Interestingly, contrary to the Ministry's claims, the consumption of subsidised LPG decreased only when there was a cap. In September 2012, the cap on number of subsidised LPG cylinders (refill) per household was limited to six per year. Later in January 2014, it was increased to nine refill cylinders. Finally, in January 2015, it was increased to 12 LPG cylinders a year.
The effect of the previous cylinder cap was also clearly visible in the composition of LPG consumption in FY2013-14, with total subsidised consumption falling in absolute terms for the first time in a decade. The unsubsidised consumption increased substantially both in absolute terms, growing by 46% relative to FY2012-13 and as a percentage of total consumption.
DBTL or Aadhaar irrelevant in blocking connections
The Ministry's latest estimate counts both inactive connections (whose status is unrelated to administrative restriction) and connections blocked over the course of several years. It must be noted here that the vast majority of such connections were identified through mechanisms that preceded direct benefit transfer (DBTL or PAHAL) and Aadhaar, and were unrelated to either initiative.
Connections that would have been valid throughout FY2014-15 were not due to introduction of the DBTL program. This is because DBTL was mandatory only in 8% of total districts for six weeks in FY2014-15. The government assumes these connections would universally have used almost double the average per connection amount of subsidized LPG, and then claims the entirety of the notional savings achieved as due to the introduction of DBTL.
In relation with Aadhaar, it must be noted that the program was always effectively irrelevant to the functioning of the DBT mechanism even for the narrow purpose of removing duplicate connections due to the household-based nature of the entitlement. The government's own figures have consistently demonstrated that the maximum number of potential duplicates identified in LPG databases through Aadhaar-based de-duplication is typically around 1% (or less) of total connections assessed, and may represent an even smaller fraction of actual consumption.
Where irregular connections were identified and blocked through the adoption of DBTL and Aadhaar in FY2014-15, a category that represents an extremely small fraction of total inactive connections recorded, any notional savings achieved would obviously only relate to the period of implementation of DBTL. Remember in FY2014-15 DBTL was only mandatory in 8% of total districts for a period of six weeks.
In other words, the vast majority of blocked connections were blocked through processes entirely unrelated to either DBTL or Aadhaar, during periods when DBTL was not being implemented, and would not have incurred any subsidy expenditure regardless of the adoption of either DBTL or Aadhaar.
Basic inquiries regarding the status and composition of the inactive connections, including the percentage of 'inactive' connections formally blocked, and within formally blocked connections, the date of identification and blocking, and the method by which they were identified for blocking, could demonstrate the extent to which this approach to estimating the fiscal effect of DBTL is incorrect or potentially highly misleading. However, it depends upon the will and wish of the Ministry to publish all such data in public domain in a transparent manner.