– do we really know who the “most polluting” industries are?

The public wants to know who is responsible for climate change, and the media has obliged with numerous exposés detailing the global greenhouse gas (GGHG) emissions of various industries.

Just this past year, the UN Committee on Climate Change posted that the fashion industry is responsible for about 8-10% of GGHG emissions. The International Panel on Climate Change’s latest publication attributes 25% of GGHG emissions to electricity and heating, and another 14% to transportation. The Guardian attributes over half of GGHG emissions to resource extraction. And just last year, Our World In Data published a report attributing 26% of GGHG emissions to food production.

But can we trust these figures – do we really know?

If you sum up the numbers above, they add up to way over 100% (128% to be exact). And in the same time, all these numbers come from more or less trustful sources.

We think we know, because reliable authorities tell us so.

But these numbers are more complex than they first appear. The reason why the statistics from early exceed 100% is because there is considerable overlap in how “industry categories” are defined. We are not being lied to, but the statistics are too vague.

In pursuit of greater transparency, Hara Kumar and Victor Castillo, two talented data scientists at KTH Royal Institute of Technology in Stockholm, Sweden, started going through the data.

– What triggered this journey was when I saw on my Facebook feed that someone had shared an article which claimed the meat industry was responsible for 25% of global emissions. Two days later I saw another article claiming that fashion is contributing 8- 10%. When I saw these two statistics together, I said, something is wrong here. Either one or both of them can’t be right”, Hara says.

Hara and Victor began by tracking down the sources behind the figures and were unable to find reputable papers supporting either claim. The reports and data they were able to find contained many disparities in their definitions and conclusions. Hara explains,

– Recently I read one of the more detailed explanation about the 8% figure for the fashion industry, however, the main purpose of the article was to report on the movement of textiles between countries and how this results in the relocation of emissions from developed to developing countries, which is a big problem with current emission tracking standards. That part of the report was well done, but when it came to measuring emissions themselves, their numbers didn’t make as much sense. Unfortunately, I didn’t have access to their data sources, so I decided to do some back of the envelop calculations to see if I could come close. In 2017, 26 million tons of cotton was manufactured and 53 million tons of polyester (according to Textile Exchange). Cotton and polyester accounts for 80% of textile industry, so I will focus only on that for now. There is 5.9 kg of CO2e per ton of spun cotton, and 9.5 kg of CO2e per ton of spun polyester (figures from Stockholm Environment Institute). In total that equates to 656 MtCO2e . There already seems to be a discrepancy, as they claim over 1400 MtCO2e for fiber and yarn production.
When it comes to energy usage, there is not really any good measure of how much energy a ton of fabric equates to. Some of the best sources I found were Carbon Trust (24kg CO2e per kg of clothing), McKinsey (20 kg CO2e per kg of clothing) and Textile World Asia (24kg CO2e per kg of clothing). The article in question, claimed over 40 kg CO2e per kg of clothing. I don’t have the data to conduct a thorough analysis myself, but these discrepancies are enough to make me doubt the validity of the 8% figure.

Among the other sources they found, estimates of the fashion industry’s share of GGHG hover around 3%, but these sources often take into account energy used by the consumer to wash, iron, and dry cloths. The media does not mention this when quoting these figures, so readers might be left imagining only the production and distribution process being responsible for X% of emissions. That is misleading, because over the lifetime of a piece of clothing, washing, drying, and ironing could consume more energy than the entire manufacturing process.

And with regards to the meat industry, Hara continues,

– The FAO places the climate impact of livestock at around 14.5% of GGHG emissions. We were not able trace the source of the 25% statistic we found on Facebook, though we suspect the number may have arisen from a report stating the entire food industry was responsible for 26% of global emissions. This illustrates how distorted numbers become by the time they finally reach the public.

These differences are not without consequences. Many people want to know what they can do to help address the climate change issue. However, simplistic answers like “no flying” or “go vegan” are not going to address the problem.

Yes, you should, if possible, take the train, instead of a plane, but even if we stop all passenger air travel, it will reduce global emissions by just 1%.

And, says Hara,

– When it comes to meat, pork and poultry emit 80% fewer emissions compared to cows, so perhaps the motto should be “no beef” instead “go vegan.”

Angry now?

Well, if these facts conflict with your deeply held beliefs, you may now feel suspicious as to whether these guys are just another source of fake news, with no other intention but to confuse.

But I can tell you, that is not the case – on the contrary!

Both Hara and Victor are deeply involved in the climate issue. Hara himself worked in the solar industry before transition to data science. However, they are concerned that the public does not have a clear understanding of the problem, who is responsible, and what needs to be done. Climate change is one of the most important issues of our time, and we cannot act or try to deal with the problem based on vague guesses (think of the European Union’s European Green Deal, where at least € 1 trillion will be spent to counteract climate change).

Instead we need a solid foundation from which we make our decisions. Because no matter what, the transformation into a fossil free society will not be without pain. So, before we press the button, let’s make sure we focus on the most relevant measures.

Here’s what Hara Kumar and Victor Castillo point out.

− The greater the distance from the point source for emissions, the greater the uncertainty about the impact of the end-user product or service on greenhouse gas emissions.
− Companies do not track embedded emissions so it is possible to relocate emissions to suppliers and distributors. For example, a manufacturer can reduce their carbon footprint by shifting energy intensive operations like die-casting to an overseas supplier, and then only assemble the components locally. By that, they can quickly claim they transitioned to 100% renewable energy because they offloaded a good portion of their energy use.
− There is great variation in the carbon footprint of industries between countries and companies. For example, American organic cotton emits only half the greenhouse gases as Indian organic cotton [according to Stockholm Environment Institute]. This makes coming up with global averages for emissions by industry or material very difficult.

To illustrate these points, Hara and Victor takes our homes as an example. Normally, when talking about the carbon footprint of homes, focus is on electricity and gas heating. Global estimates vary, but residential energy use accounts for 9% of global emissions. However, it takes energy to mine the coal and drill the oil and gas needed to meet residential energy needs. The extraction and processing release some GHGs directly. This is typically treated as a separate category in climate change statistics. Those arguing for full supply chain accounting will allocate these emissions to the end users of energy. Doing that will increase residential energy emissions to around 12-13%.

But that part is just the energy the resident is using. What about,

– the emissions coming from making the concrete, steel, and lumber needed to construct the building? Approximately 7-12%.
– the energy used in the construction process? Approximately 1-3%
– the furniture and appliances inside the home? Approximately 9-10% (but highly uncertain).
– the leakage of refrigerants from AC units? Approximately 1-2%

Add up all that and suddenly the emissions of your home jumps to between 30% and 40%.

If you think, ‘wait a minute there is a huge difference between 30% and 40%’, you are right, and it perfectly illustrates the difficulty of measuring emissions as you travel further and further from point source of emissions.

We can say with a good degree of confidence how much energy homes use, and how much GGHG emissions can be attributed to cement / concrete. But how does that ultimately reach the consumer?

Lumping all this into a category called “industry” does nothing to inform consumers or policy makers where they should focus their attention. And that leads to poor investment decisions.

For example, retrofitting a building with better insulation will reduce energy consumption and traditional methods of accounting for emissions would applaud such an action every single time. However, we must also take into account the energy used to manufacture, transport, and install the new insulation. If the existing insulation is good enough, it might even be climate smart to not invest in the upgrade, and focus efforts somewhere else.

Governments must require companies to track not just their direct emissions, but emissions embedded in their entire supply chain. When making investments to reduce emissions, the emissions embedded in retrofits and new equipment must also be accounted for, in the same way financial accounting uses depreciation to allocate long-term investments into cost of goods sold.

– Without more rigorous accounting, government polices like a carbon tax may actually encourage companies to make environmental investments that appear to nominally reduce emissions but in reality do not, or possibly even cause more damage.
says Hara and Victor.

It also encourages companies to engage in practices like off-shoring and subcontracting, which only further blur our data and obstruct our ability to tackle the climate problem.

Hara Kumar and Victor Castillo are not satisfied that existing tools and methods are up to the challenge. So now they have started building their own software solution called Emission Atlas, which aims to help companies get more accurate figures regarding their climate impact.

It builds off of existing company systems, like ERPs (Enterprise Resource Planning System), and in addition to calculating a company’s wide emissions footprint, the tool will calculate the emission efficiency of each value chain.

Because GHG emissions are a reasonable proxy for the amount of energy and material put into a process, comparing GHG emissions between two comparable processes can reveal inefficiencies before they appear on financial statements. So beside getting the opportunity to lower the GHG footprint, Emission Atlas also gives the company the possibility to trace and lower the cost for production.

The next step for Hara and Victor is to find a company in Sweden to work with and use real data as they build their software. They are focusing on industries involved in housing: furniture, flooring, insulation, appliances, etc. However, they are open to working with manufacturers in the food and fashion industries as well, given the industry’s recent efforts to reduce their GHG footprint.

I guess IKEA, HM and others have reason to expect a call within the coming period.

Listen to my latest video blog with Hara Kumar, Founder and CEO at Emission Atlas, giving the full story behind the concept.

Also, find more about Emission Atlas at

See ‘ya!

Leave a comment