Using research to assess extent of
‘Climate injustice’ in India
A research paper
Prepared by: IMRS Advisory, Bangalore
1
The context
The context was of Kyoto protocol, countries around the world debating to
agree on commitments for larger carbon emission cuts to keep rising global
temperatures in control. Government of India demanded ‘common but
differentiated’ responsibility w.r.t carbon emission. Developed nations need
to cut their CO2 emissions not only to prevent climate change but also to
allow developing world to catch up, without pushing the global temperatures
beyond the crisis level.
Greenpeace, an international environment related NGO, asked a critical
question – is there ‘climate justice’ in India? They wished to present a case to
the Indian government to implement the principle of ‘common but
differentiated responsibilities’ across the various economic classes within the
country. They believed that there are critical linkages between development
and environment sustainability, and that the rich contribute significantly to
environmental pollution. They believed that at a per capita level, the rich in
India contribute to as much carbon emission as do the developed countries.
They wished to make a point on ‘climate injustice’ in India and therefore
‘common but differentiated climate responsibilities’ of different socio
economic segments, with the help of a market research exercise conducted
across the country and across various socio economic classes.
This was broadly the given context in which this research was done.
The approach and the challenges
The key challenge in this brief was the fact that this was quite unlike any
market research brief ever taken by us. This was in a category we would
have never considered ‘researchable’. Additionally, the findings had to be
presented by Greenpeace to make their point before Government, public and
media and therefore had to be impeccable in the broad approach and all finer
details.
Therefore, in consultation with the client, it was agreed upon that we would
need to do a first of its kind, face to face survey across the country ranging
from the metros, medium to small towns and rural areas on domestic energy
2
consumption and transportation, covering both medium used and quantity for
the same. Primarily, we would need to enlist all items of domestic energy use,
confirm the type of energy source used, assess heaviness of usage and
convert that to carbon emission with norms already established.
Domestic carbon emission was to be estimated from the following sources:
• Electricity consumption – appliance and capacity; in rural, pump set for
field should be out of calculation
• Cooking Fuel consumption
– Fire wood, gober and gober gas considered as renewable
sources and not added to CO2 emission
• Conveyance - Petrol / diesel consumption, flights taken
So the approach sounded simple - assess incidence and heaviness of the
above mentioned and convert using norms to carbon emission.
Key challenges on the project, post arriving upon the broad approach were
manifold:
1) Who should be interviewed to represent the household? – Post
some pilots we figured that it would have to be a mix of person
spending maximum time at home and also CWE. Person spending
most time in the Household (not necessarily housewife or CWE) is best
equipped to answer questions regarding heaviness of electricity
consumption etc.
2) Can people give exact estimate of how many hours each electric
bulb / fan works in their house? Difficult, not impossible. An
alternative was to take electricity bill, but not all have bills, nor do all
have a clear idea of bill amount in summer vs. winter etc. So we
decided to follow a dual approach, we took details of each electric item
and its consumption in the household and also validated this with
electric bill estimates.
3) How to factor in seasonality as bill amounts can vary by season?
Respondents were asked to give their estimated bill amount for winter
months, summer months and rest of year. They were also asked to
give their estimates of how many months would they classify as winter
months, summer months and rest of year. Weighted average of
seasonal bill with length of season (No. of months) was done to arrive
upon annual electricity bill of the household.
3
4) How to convert appliance related details to total electricity
consumption? We captured data on each electric appliance at
household level - appliances used, nos. used, months used in
(especially for seasonal appliances like heater / gyser etc), days used
in a month, and no. of hours used per day. For electric bulbs, CFL, tube
light, heater, gyser etc, watts were also recorded. For appliances like
TV & refrigerator, same size of appliance can have different electric
consumption for different brands and models and therefore, brand and
model details were taken from HH and secondary estimates of their
wattages were used. Thus, for all appliances, total electricity
consumption was calculated. Comparison of this with estimated
electricity consumption showed that the two estimates were in line
(variance was within 5% error band); in few cases, these two had a
wide variance, these cases were rejected from analysis. This appliance
analysis was also used to assess contribution of different appliances to
the total electricity consumption and therefore, total carbon emission.
5) How to conveyance fuel consumption estimates – It was easy to
estimate consumption of petrol / diesel by 2 wheelers and 4 wheelers
owned by the Household as the CWE would have estimates of how
often and how much, they fill their vehicles. However, it became tricky
when it came to usage of public transport. We took distance traveled
estimates from the respondents and used other primary / secondary
estimates to assess their fuel consumption estimates per km traveled
to arrive upon consumption value of each individual.
There were many such issues which made the research a logistical
nightmare. Post a pilot, we arrived upon solutions to all such issues and
norms to be used for carbon emission conversion. Final methodology used
for carbon emission estimation is detailed herewith.
Methodology for carbon emission calculation
A) Electricity consumption
Step 1 – take the total monthly bill by seasons, weight by length of season
and arrive upon annual bill
Step 2 - convert the bill to units, using the system followed by the electricity
board
4
Step 3 – validate with appliances used and details of usage as explained
above and arrive upon total units consumed
Step 4 – reject cases where variance between the 2 methods was more than
20%
Step 5 – convert all electric units to carbon estimates
B) Cooking fuel consumption
Step 1 – estimate the different fuels used, excluding electric as it is already
covered and excluding renewable resources and estimate weight of fuel
consumed
Step 2 – convert all fuel weight to carbon estimates
C) Conveyance fuel consumption
Step 1 – estimate fuel consumed by 2wheeler / 4 wheeler owned
Step 2 - estimate distances traveled on public transports, including air travel
Step 3 – distance converted to fuel weight
Step 4 – convert all fuel weight to carbon estimates
Add all the estimates above to arrive upon total estimates for the family and
there-from calculate per capita carbon emission.
Caveat: This study aimed at estimating only carbon footprint emerging from
personal domestic energy consumption. Commercial energy consumption in
production of goods etc was not within its ambit. Therefore, any electricity /
fuel consumption for commercial purposes was kept out of calculations even
if the usage was happening within household. Examples of these are – water
pump used for irrigation was kept out but electrical pump for home water was
considered as personal energy consumption.
Key findings
The findings proved the client’s hypothesis and showed that a significant
carbon footprint of India is contributed by a relatively small wealthy class.
Greenpeace made a case of ‘climate injustice’ in India. They also shared this
with the media and their report was published on their site as a whitepaper –
‘Hiding behind the poor’.
Some of the key findings of this study are shown herewith:
5
1.1. CO2 emission per household by different segments
Much higher CO2 emission observed in metros and in higher income
segments – direct correlation can be seen between CO2 emission and pop
strata / income. Average Carbon footprint of the highest income segment is
almost 4 times that of the lowest income segment.
6000 Fig 1 - Kgs of CO2 emitted per annum by average HH
5000
4000
3000 5433
2000 3654
3010 2975 3151
2382 2435 2561
1000 1313 1617
1169
0
All Metro Non- Rural < 3k 3-5k 5-8k 8-10k 10-15k 15-30k 30k+
Metro
1.2. Contribution of different energy sources in CO2 emission for
different income classes
The largest contributor to this carbon imprint is electricity, which contributes
to about 2/3rd of all carbon emissions. % contribution of electricity increases
as incomes increase - this is 59% in the lowest income group, going up to
73% in the highest income segment. Transportation % also increases with
income levels, while % contribution of cooking fuel drops.
Fig 2 – Contribution to carbon emissions
100% 1600
14 14 12 12
90% 16 18 15 16 21 17 19 1494
1400
80%
21 18 21 29 28 20 18 14 8 1200
70% 24 15
843 827 1000
60%
936
50% 685 800
501 633 819
40% 600
67 69 73
65 66 65 60 465 65 63
30% 58 372 59
400
20% 335
10% 200
0% 0
All Metro Non-Metro Rural < 3k 3-5k 5-8k 8-10k 10-15k 15-30k 30k+
% of Electricity % of Cooking Fuel
% of Vehicle / Transportation Per capita Kgs CO2 emission (per annum)
6
1.3. Climate Injustice Quotient – across income segments in India
The concept of Climate injustice is simple – proportion of carbon emission of
any segment as a ratio to the proportion of the population of that segment.
The table below shows the inequality in carbon emission for different
segments. Segments with CIQ of less than 1 are emitting less than average
and the highest income segment here is emitting 3 times the average.
Table 1 – CIQ across income segments in India
Income Population Share of Annual CO2 Share of CIQ
classes (2007) population emission in Emissions
billions
All 112986615 100 566 100 1.000
<3k 432162957 38 145 26 0.668
3-5k 390796483 35 182 32 0.926
5-8k 155730946 14 107 19 1.365
8-10k 69178755 6 57 10 1.643
10-15k 53236409 5 44 8 1.652
15-30k 18804012 2 18 3 1.910
30k+ 9956500 1 15 3 3.010
Graph below is a visual depiction of CIQ (Carbon Injustice Quotient) across
income classes in India.
Fig 3 – CIQ across income segments in India
3.01
1.91
1.643 1.652
1.4. International1.365
Climate Injustice
That brings
.668 us to
.926 the ultimate question of Climate injustice in the world and
how <the3k principle
3-5k of ‘common
5-8k but8-10kdifferentiated
10-15k responsibility’
15-30k cannot
30k+ be
applied to world without also being applied to India.
For this, first need is to make the findings of this study comparable to
international studies. India’s overall per capita CO2 emission is estimated at
1.67 tons by World Research Institute, while the annual per capita CO2
emission estimated by this study (personal domestic carbon emission) is 501
7
kg. This is 33% of the total and is inline with the sectoral division of CO2
assessed by WRI. To enable comparison of per capita emissions of income
classes with national CO2 emission of other countries, the emissions were
multiplied by a factor of 3.3 to ensure that total Indian emissions
corresponded to an overall of 1.67 tons per capita.
Using this analysis, per capita emission and CIQ of income classes from
India have been plotted on the international emission table given herewith.
Table 2 – CIQ in the world
Global Per capita
Share of emissions Share of CO2
World in Mln Global emission
Country / Income Class population tons Emissions CIQ (in Tons)
USA 4.45% 6871.7 20.38 4.58 23.06
Russia 2.12% 1915.7 5.68 2.68 13.49
Germany 1.22% 1013.3 3.01 2.46 12.36
UK 0.90% 658.8 1.95 2.18 10.98
Japan 1.90% 1351.5 4.01 2.12 10.64
France 0.90% 518.4 1.54 1.72 8.64
World 100.00% 33733.93 100 1 5.03
India (Monthly Income: Rs. 30k) 0.15% 49.52 0.15 0.99 4.97
Brazil 2.81% 849.5 2.52 0.9 4.52
China 19.55% 4963.1 14.72 0.75 3.79
India (Monthly Income: Rs. 15 to 30k) 0.28% 58.61 0.17 0.62 3.12
India (Monthly Income: Rs. 10 to 15k) 0.79% 146.62 0.43 0.55 2.75
India (Monthly Income: Rs. 8 to 10k) 1.03% 188.68 0.56 0.54 2.73
Sustainable Average Per Capita CO2 Emission 2.5
India (Monthly Income: Rs. 5 to 8k) 2.32% 355.24 1.05 0.45 2.28
Indonesia 3.66% 504.6 1.5 0.41 2.06
India 16.86% 1889.1 5.6 0.33 1.67
India (Monthly Income: Rs. 3 to 5k) 5.83% 605.13 1.79 0.31 1.55
India (Monthly Income: < Rs. 3k) 6.45% 481.85 1.43 0.22 1.11
Bangladesh 2.19% 122.4 0.36 0.16 0.83
The highest emission segment in India is less than the World average and
has a CIQ of less than 1. However, sustainable development needs per
capita carbon emission of 2.5 tons per annum. In India there are 4 income
segments ie., about 150 mln people earning more than Rs. 8000 pm, who
emit more than 2.5 tons of CO2 per capita per annum. For the balance 980
mln people to hope for development, and increase their carbon
8
footprint, the top 150 mln will have to consider ways and means to
reduce their carbon footprint.
Learnings for research
There were several learnings for us in the process of conducting this
research. The most interesting learning for us was to experience using
research in the hitherto unexplored category like environmental studies. If
used rigorously in conjunction with other scientific secondary learnings,
research can be used productively in sectors which might be perceived
‘unresearchable’ and significantly add to our social and environmental
knowledge and understanding.
Another key learning in research context was that interviewing the CWE or
HW alone may not be sufficient to map household consumption in many
categories – this arena is the forte of the person who stays at home for the
maximum time, who at times might not be the HW, especially in case of
households with a working HW. Person spending maximum time at home can
be an interesting concept and can be adopted as the main respondent in HH
consumption mapping studies. Secondly, in most consumption mapping
studies, we normally do not tend to explore and differentiate between
household consumption & household production intended for commercial
usage. This study starkly highlights that we may need to re-look at some
taken-for-granted assumptions prior to approaching studies that map
household consumption.
---X---X---X---