Copenhagen Consensus Center

Report Part Title: Free provison of LPG connection to poor households

Report Title: Benefits and Costs of Household Air Pollution Control Interventions in
Rajasthan
Report Author(s): Bjorn Larsen
Copenhagen Consensus Center (2018)

Stable URL: http://www.jstor.com/stable/resrep21836.9

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Copenhagen Consensus Center is collaborating with JSTOR to digitize, preserve and extend
access to this content.

This content downloaded from

218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
 Table 4.6. Quality of evidence
Quality of evidence
Baseline health data Medium-Strong
Relative risks (RR) for health benefits Medium-Strong
Valuation of mortality Medium-Strong
Sustained user rate of intervention Limited-Medium
Time savings Medium
Cost of intervention Medium-Strong
Total evidence Medium
Source: Author.

5. Free provison of LPG connection to poor households

5.1 Description of intervention

While over 80% of urban households use clean fuels (mainly LPG) for cooking, only 15% do so
in rural areas according to the NFHS IV 2015-16 (IIPS, 2017). Nationwide 24% of rural
households use clean cooking fuels. The rates of clean cooking fuel utilization are even lower
among the poorer segments of the population.

The government has therefore implemented a program (Pradhan Mantri Ujjwala Yojana
(PMUY) launched in 2016) that provides free LPG connections (LPG cylinder and auxiliary
equipment) free of charge to households below the poverty line (BPL) to encourage these
households to switch from solid fuels to LPG. The budgeted cost to the government is Rs.
1,600 per connection. The households can also get a loan from the oil marketing companies
to cover the cost of an LPG stove.8 A loan can also be obtained for the first LPG filling of the
cylinder.9

The intervention assessed in this paper, in terms of benefits and costs, is therefore the free
provision of LPG connection to poor households.

5.2 Intervention response rate

Studies of household adoption of improved cookstoves and clean fuels have identified
upfront cost as a major obstacle (Lewis et al, 2015), as discussed under the ICS intervention.

27
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
Thus free provision of one of the cost components of cooking with LPG may be expected to
induce some households to switch to LPG. However, the LPG stove is also an important cost
component. An LPG stove with two burners costs about the same as the connection.
Moreover, LPG fuel even at current subsidized prices in India costs several times more per
year than the LPG connection.

Important questions are therefore to what extent free provision of LPG connection induces
households to switch to LPG, how much of cooking will be done with LPG, and how sustained
is the switch to LPG.

A survey undertaken by financial consulting firm Micro Save in 12 districts of Uttar Pradesh,
India revealed that nearly all of the beneficiaries of the scheme switched to cooking with LPG
as soon as the LPG cylinders were made available.10

However, a large number of LPG connection beneficiaries have not come back for refills in
many states. The gap between the growth in LPG connections and LPG consumption in 2016-
17 confirms the ground-based reporting of PMUY customers not buying refills.11

Moreover, a survey from Uttar Pradesh revealed that refilling LPG cylinder was done only
four times in the past one year by some beneficiaries (approx. 1/3rd of total energy need for
cooking). This was primarily because the beneficiary finds refilling too costly.12

Data on annual LPG consumption and LPG connections can shed some light on the adoption
rate of LPG, or response rate, among BPL households resulting from free LPG connections.
New connections increased by 32.2 million in 2016-17, of which 20 million were PMUY
customers that received free connections.13 Assuming that households with LPG connection
in 2015-16 (prior to PMUY program), as well as the new non-PMUY connections in 2016-17,
consumed the same amount of LPG (kg/household) in 2016-17 as they did in 2015-16 would
imply that the new PMUY connections consumed on average about 55 kg of LPG per year
(about 55% of average non-PMUY consumption per household per year). This is an average
and one can expect that some households adopted LPG as their primary cooking fuel while
others used LPG only for certain cooking needs (or even eventually abanonded the use of
LPG).

28
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
Estimating the number of PMUY connections that adopt LPG as primary cooking fuel is of
main interest because it is these households that will achieve the most substantial health
benefits. Use of LPG for let’s say only 20-25% of a household’s cooking needs would be
expected to result in only modest reductions in household members’ PM2.5 exposure, and
thus very modest health benefits.

Based on reported household consumption of biomass fuel in Rajasthan (Nielsen India, 2016)
one can estimate that a household would need about 150 kg of LPG per year to meet its
cooking energy needs if it were to use LPG as exclusive or primary cooking fuel. PMUY
households that use LPG as a secondary fuel may on average consume 15-30 kg per year, or
about one to two cylinders.

If one applies a range of 100-150 kg of LPG consumption for households that use LPG as
primary cooking fuel, and a range of 15-30 kg for secondary users, one can estimate that
around 21-48% of BPL households receiving free LPG connections adopt LPG as a primary
cooking fuel. The mid-point of this range is 35% and is used as the household LPG adoption
response rate to free LPG connections in this paper.

5.3 Calculation of Costs and Benefits

5.3.1 Costs
Costs associated with the intervention are the government provided LPG connection (cylinder
and auxiliary equipment), as well as LPG stove, stove maintenance and repair (O&M), and
LPG fuel.

The government budgeted cost of LPG connection is Rs. 1,600 per household. This implies an
effective cost of Rs. 4,570 per household, assuming that 35% of the households adopt LPG as
primary cooking fuel as previously discussed.

Households that receive free LPG connection will also purchase LPG stove. Cost of a two-
burner LPG stove is also about Rs. 1,600.14 However, only 35% of the households (LPG used
as primary cooking fuel) will receive sustained and substantial benefits from LPG. Thus the
effective cost is again Rs. 4,570 per household.

29
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
 Annual O&M is assumed to be 5% of stove cost, or Rs. 80 per year. LPG fuel cost among
households using LPG as primary fuel is estimated at about Rs. 7,077 per year based on a
consumption of 150 kg per year and a price of Rs. 670 per bottle (14.2 kg). This was the
average 10 months non-subsidized market price from August 1st 2017 to April 1st 2018 in
major markets in India.15 The market price, and not subsidized price, is used to estimate cost
and benefits of interventions because both private and public costs shall be included in the
assessment.

Annualized cost per household is estimated at Rs. 8,285 (table 5.1) and total annualized cost
of intervention is estimated at Rs. 14.5 billion (table 5.2) based on total intervention
beneficiaries of 1.75 million households, i.e., households adopting LPG as primary cooking
fuel.

Table 5.1. Cost of intervention (Rs per household)

Initial cost Effective cost Annualized cost
Cost of connection 1,600 4,570 564*
Cost of stove 1,600 4,570 564*
O&M (5% of stove cost per year) 80
LPG fuel cost 7,077
Total annualized intervention cost 8,285
Note: Annualized cost is calculated using a discount rate of 5%. * Useful life is 10 years.
Table 5.2. Total annualized cost of intervention, Rs million
Total
Beneficiary households (000) 1,750
Total annualized cost, Rs million 14,499
Note: Discount rate: 5%. Source: Estimates by author.

5.3.2 Benefits
Health benefits
Health benefits of moving from pre-intervention to post-intervention exposure levels for
intervention households that adopt LPG as primar cooking fuel are estimated by using the
integrated exposure-response (IER) methodology from the GBD 2015 Project presented in
annex 1.

Estimated percentage reduction in health effects among beneficiary households is 45% if the
households consistently use LPG as primary fuel. This relatively low percentage reduction in
health effects is due to the post-intervention PM2.5 exposure of 50 µg/m3 among adult
women and children and 35 µg/m3 among adult men. These relatively high exposure levels

30
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
are associated with air pollution from surrounding households that continue to use solid fuels
for cooking, as well as from the use of solid fuels as secondary cooking fuels in the household
that uses LPG as primary fuel.

The estimated reduction in health effects from the intervention amounts to 4,956 deaths
averted and 9,112 YLDs per year (table 5.3).

Table 5.3. Health benefits of intervention

Averted deaths per year 4,956
Averted YLDs per year 9,112
Source: Estimates by author.

Switching to LPG also has non-health benefits. Main benefits are reduced biomass fuel
consumption, whether self-collected or purchased, and reduced cooking time. The
magnitude of these benefits will depend on current cooking arrangements, household
cooking patterns, cost of fuels, and household member valuation of time savings.

Fuel savings
Estimation of the value of solid fuel savings from switching to LPG as primary cooking fuel
follows the method applied under the improved cookstove intervention.

Households in the state spend on average 23 hours per month on solid fuel collection and
preparation (Nielsen India, 2016). The value of time savings associated with no longer having
to undertake this activity can be estimated based female wages rate, and a value of time
equal to 50% of the female wage rate. Thus the annualized value of time savings amount to
Rs. 7,073 per household per year over the lifetime of the LPG stove (i.e., 10 years) at 5%
discount rate. A female wage rate is applied as most fuel collection and preparation is
carried out by women.

Cooking time savings

PAC (2014) reports an average cooking time saving of nearly 30% from the use of an LPG
stove in South Asia. Average cooking time in Rajasthan in a household using solid fuels and a
traditional cookstove is 142 minutes per day (Nielsen India, 2016). Thus cooking time savings
are estimated at 43 minutes per day. A value of time equal to 50% of female wage rates is
applied to estimate the value of cooking time savings. Annualized value of time savings over
the life of the improved cookstove (at 5% discount rate) is Rs. 6,551 per household per year.

31
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
 Total benefits
The total annualized value of benefits of the intervention is estimated at Rs. 42 – 70 billion,
depending on method used for valuation of deaths averted, i.e., VSL or YLL at 3 times GDP
per capita (YLDs are valued at 3 times GDP per capita). Annualized benefits per household
that adopts LPG as primary cooking fuel are Rs. 26,937 when averted deaths are valued using
YLLs valued at 3 times GDP per capita, and Rs. 39,180 when averted deaths are valued using
VSL. This includes CO2 emisison benefits presented in section 3.3 (table 5.4).16

Table 5.4. Value of benefits of intervention, Rs per household per year

VSL+YLD YLL+YLD
Health benefits 25,408 13,165
Fuel collection time savings 7,073 7,073
Cooking time savings 6,551 6,551
CO2 emission benefits 149 149
Total benefits 39,180 26,937
Note: Discount rate: 5%. Source: Estimates by author.

3.3.3 Benefit-cost ratios

A comparison of benefits and costs, and benefit-cost ratios (BCRs) are presented in table 5.5.
BCRs are in the range of 4.5-4.9 when averted deaths are valued using VSL and in the range
of 2.8-36 when averted deaths are valued using YLL at 3 times GDP per capita.

Table 5.5 Benefits and costs of intervention, Rs million per year and BCRs
3% discount rate 5% discount rate 8% discount rate
Valuation method Benefit Cost BCR Benefit Cost BCR Benefit Cost BCR
VSL+YLD 70,231 14,347 4.9 68,566 14,499 4.7 66,748 14,733 4.5
YLL+YLD 52,278 14,347 3.6 47,140 14,499 3.3 41,778 14,733 2.8
Source: Estimates by author.

5.4 Assessment of Quality of Evidence

The dimensions that most importantly affect the estimated benefits and costs of the
intervention are presented in table 5.6. Quantified health benefits of the intervention are
proportional to the baseline health data. These data are most likely of medium-strong
quality. The relative risks (RR) of disease and mortality reductions are based on a large body
of global research, but not specifically in Rajasthan. The value of statistical life (VSL) used for
valuation of mortality benefits is from a benefit-transfer function developed by the World
Bank (World Bank, 2016). The function is based on meta-analysis of VSL studies from mostly
high- and medium-income countries and other available evidence of VSL by country income

32
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms
level. The rate of sustained adoption of the intervention has limited-medium evidence but
does not substantially affect the BCRs. The time savings are based on medium evidence as
studies from Rajasthan are limited.

Cost of intervention has strong evidence.

Table 5.6. Quality of evidence

Quality of evidence
Baseline health data Medium-Strong
Relative risks (RR) for health benefits Medium-Strong
Valuation of mortality Medium-Strong
Sustained adoption rate of intervention Limited-Medium
Time savings Medium
Cost of intervention Strong
Total evidence Medium-Strong
Source: Author.

6. Reduction of subsidies to LPG fuel

6.1 Description of intervention

LPG fuel has long been subsidized or priced below market price in India. The government
moved towards closing the gap by gradually increasing the subsidized price. As of August 1st
2017 the average subsidy in four major urban markets was reduced to less than 10% of non-
subsidized market price. However, world prices of crude oil and LPG have since increased.
As of April 1st 2018 the non-subsidized price of LPG had increased by 24% since August 1st
2017 while the subsidized price increased by 2%. Consequently the average subsidy in these
four markets was 25% of non-subsidized market price. The subsidy rate peaked at about 33%
during November 2017 to February 2018 when the non-subsidized price was highest.

Increasing the subsidized price of LPG to reduce or eliminate the LPG fuel subsidy is likely to
make some households cut LPG consumption and increase the use of solid fuels for cooking.
This entails negative health effects. On the other hand, subsidy reduction will reduce the
resource allocation inefficiency that subsidies create, simplest measured by the so-called
deadweight loss.

33
This content downloaded from
218.248.44.196 on Tue, 26 Dec 2023 08:25:02 +00:00
All use subject to https://about.jstor.org/terms