The Case for CATs (Carbon Added Taxes) –

Tackling Environmental Externalities in a Global Economy

By Salman Farmanfarmaian^*

  salmanff.blogspot.com

A Carbon Added Tax (CAT) is like a sales tax on the carbon emitted to produce the goods and services we consume. From the CAT perspective, when you buy say, a computer, you are responsible for the carbon emitted in producing its mother board and the hard disk and each subcomponent in the computer, as well as that emitted from the fuel burnt by ships and trucks transporting it, and to power the electricity in the shop selling it. A CAT levies a tax on the consumer for all this “embodied” carbon, at the point of sale.

This paper argues that a CAT represents a better way to affect global carbon emissions, because it leverages the global nature of the world economy to cut through geographic and legislative boundaries. The paper also addresses the major arguments against the CAT – the assumed difficulty of measuring embodied carbon and administering the tax. It discusses the latest developments in this area, and proposes structural incentives to overcome those obstacles.

The Moral Imperative: Consumer vs Producer Responsibility¹

Over the past decade or so, a number of regulatory initiatives have aimed to curb green house gas emissions. But these efforts have largely tackled the problem from a “producer-responsibility” perspective. That is, nations and states have tried to impose limitations or penalties on the carbon emissions emanating from within their jurisdictions. For example, the European Trading Scheme (ETS)² has largely tackled carbon intensive industries within the European Union; proposed legislation in the US has also focused on industries within the states’ borders.

The Consumer-Responsibility paradigm would suggest a different approach. The premise is that carbon is emitted to produce all of the products and services we consume – all the carbon “embodied” in the products we buy. And as consumers, we are responsible for causing the emissions through out their production process, right on to the end of their lives.

As such, the primary case for such a Carbon Added Tax is one of moral equivalency. If, as a society, we care about the carbon we are responsible for emitting, then we should care as much about the carbon we emit directly in our jurisdictions, as for the carbon emitted outside of our borders to produce the goods that we consume at home.

For example, if we want to impose a carbon tax on the emissions of local cement factory, there is no reason that the same tax should not be imposed on cement that is produced outside of our jurisdiction, but which is imported for local consumption. After all, it is the consumer of that cement which is creating the demand for the cement, wherever it may be produced.

A Word on Leakage

Most discussions about carbon emissions outside of regulated jurisdictions focus on the idea of “leakage” – a fundamentally producer-focused question of whether carbon regulations in one jurisdiction would simply push the emissions outside of those borders without affecting the overall carbon emissions in the world. For example, it has been argued that a tax on carbon in Europe might force carbon-intensive industries such as cement to produce the same products outside of the European Union, and subsequently transport it to the EU, presumably emitting more carbon per ton of cement than in the absence of such regulation.³ Other studies, often conducted by proponents of carbon regulations, have argued that the impact of leakage on industry is non-material.⁴ But material or not, it is the producer-focused paradigm of the leakage debate which is flawed. From a consumer-responsibility perspective, it does not matter where the carbon was emitted, much less whether the emissions migrated to a new geography. What matters, is the reason for which that carbon was emitted: to produce a good and deliver it to a consumer. As such, it would only be logical to reduce carbon emissions by tackling it at the point of consumption.

The Quantitative Impact of the Consumer Responsibility Perspective

Looking at carbon emissions from the Consumer responsibility perspective can change the picture of environmental footprints quite radically. For example, a high level analysis of embodied carbon in imports by Oxford’s Dieter Helm⁵ looks at the UK’s carbon emissions from the Consumption perspective, and notes that using conventional producer-based carbon accounting-methods, “UK greenhouse gas emissions have fallen by 15% since 1990. In contrast, on a consumption basis, the illustrative outcome is a rise in emissions of 19% over the same period. This is a dramatic reversal of fortune… It suggests that the decline in greenhouse gas emissions from the UK economy may have been to a considerable degree an illusion.”

Another study from Carnegie Mellon uses a Life Cycle Analysis Input Output database to estimate the total amount of carbon embodied in products imported into the US to be between 0.8 and 1.8 Gt in 2004 – an average of 1.2 Gt. To put this number in perspective, that would be more than all the carbon emitted in US homes that same year (including all residential electricity and heating.)

Using the same basic dataset, a study by Bin and Dowlatabadi⁶ estimated that less than 50% of the emissions related to consumer expenditures are due to direct influences, and more than half would be due to indirect influences – in other words, “embodied” in the goods and services consumers buy.

All these studies show that embodied emissions represent a vey large portion of the total carbon emissions each of us is responsible for. That is, from a consumer responsibility perspective, our “carbon footprint” is actually much larger than what we often imagine.

The Economics: A Global Public Good

Carbon emissions are fundamentally global in nature, and in more than one way. First, the world’s climate is the quintessential public good, and so green house gas emissions create the quintessential global negative externality⁷: A molecule of carbon dioxide emitted in Vietnam has as much effect on global warming as a molecule of CO2 released in Munich.

Second, the majority of the increase in emissions of green house gases over the next decades is projected to take place within developing nations. According to the IEA’s reference scenario projections, three quarters of new energy capacity, and 97% of the increase in energy-related CO2 until 2030 will arise in developing countries.⁸ That is, based on current and expected CO2 regulation in the developed world, only 3% of the increase in CO2 emissions in the next 20 years will be coming from there. So any further effort towards curbing global emissions must tackle the problem globally. To continue to focus on regulating carbon emissions in the developed world alone, is to miss the crux of the problem.

Third, an often overlooked point, the increasingly global nature of the world economy means that the products produced in those developing countries often end up in the more developed economies of Europe and America for consumption. That molecule of carbon dioxide emitted in Vietnam may well have been emitted to produce something that would be exported to Germany and consumed by someone in Munich. (In other words, that molecule may well have been embodied in a product sold in Germany.) So if we focus on local emissions in developed countries, at best we would be worrying about an immaterial problem, and at best, whitewashing our own responsibility as a consumer of the goods produced in developing nations.

Structural Problems with the Producer Responsibility Paradigm

International efforts to curb global warming, such as the Kyoto agreement and its potential successor, have concentrated on inter-governmental agreements to regulate the level of carbon emissions “produced” in each country. This producer focused effort has led to some fundamental disagreements among nations. On the one hand, the more advanced economies such as the United States and Europe have been striving to reduce actual emissions per country or per GDP. On the other hand, developing economies like China argue that their emissions per capita are currently far lower than those of advanced economies, and that they need to follow the same rules the Western World did, when it went through the industrial revolution and adopted a more “advanced” economy. As such, they argue that imposing restrictions on them would be unfair. Within a producer-responsibility paradigm, both sides of the argument are absolutely right, thus creating the impasse we have witnessed since the Kyoto agreement more than a decade ago.

Developing countries also note that economies naturally go through a cycle of development. As they mature, they migrate away from manufacturing industries and become more and more service oriented, and thus emit less carbon per dollar of GDP. Yet, this cycle does not mean that these countries will be consuming less goods – just that they will be producing less goods locally. A new generation of up and coming developing nations inevitably rises to provide the manufacturing services world consumption demands of them. And in doing so, these newly developing countries might make the same arguments for why they should be creating the next large coal fired power plant to fuel their growth. Mean while, the world as a whole would be consuming more and causing more emissions globally.

So, within the producer responsibility paradigm, the global nature of the world economy is an intrinsic part of the problem: Just as advanced service oriented economies instill carbon taxes and solar panels, and a false sense of security in their producer based emissions metrics, production of energy intensive goods in our global economy shifts to new developing economies which can continue to evade international agreements on (producer based) carbon emissions. And as developing nations produce the carbon intensive goods that are consumed in the developed world, they contribute to a cycle that many developed-nation consumers are trying to stop.

In contrast, the consumer responsibility paradigm takes advantage of the global nature of today’s world economy. It leverages the purchasing power of the world’s richest consumers, to affect carbon emissions through out the supply chain. Kyoto style international agreements use the command and control levers of inter-governmental agreements to try and regulate an inherently distributed and market oriented global supply chain – an effort which is inherently inefficient. In contrast, the consumer-responsibility paradigm introduces a distributed mechanism coherent with the global nature of today’s economy, to create a market oriented harmonization of carbon taxes.⁹

A Carbon Added Tax (CAT) takes precisely such a consumer-responsibility approach to regulating carbon emissions. How would that work in practice?

The CAT in Practice

In a recent speech, Nobel laureate Joseph Stiglitz outlined a proposal for imposing a CAT:

“A carbon added tax (CAT), levied at each stage of production, would have some of the same advantages that a value added consumption tax has. Each producer would have to show receipts for the carbon tax paid on inputs into its production. The taxes levied at each stage of production would be passed on to consumers. It is as if the tax were imposed on consumers… A carbon value added tax will both discourage production in more carbon intensive ways and discourage the consumption of carbon intensive goods…

“Thus, while it might seem administratively simpler to impose the tax at the point of production of coal, oil, or gas, or at the cutting of the forest, etc., any carbon tax system will have to focus on usage, i.e. imposing the tax on the use of carbon (oil, coal, gas) at each stage of production.”¹⁰

Opponents of taxing embodied carbon argue that imposing such a tax is impractical. The objections can be split into two categories. The first is that such a tax would be difficult to administer, and the second, that calculating the embodied emissions of products would be, to use the words of the Economist, an “impossibly complicated task.”¹¹

In regards to the administrative issue, it is important to note that a CAT would be administratively similar to the VAT which has been implemented successfully in much of Europe and the world. In fact, the mechanism outlined by Stiglitz mimics almost exactly the process through which the VAT is collected in Europe, as outlined by the EU’s own web site¹²:

“The VAT due on any sale is a percentage of the sale price but from this the taxable person is entitled to deduct all the tax already paid at the preceding stage. Therefore, double taxation is avoided and tax is paid only on the value added at each stage of production and distribution. In this way, as the final price of the product is equal to the sum of the values added at each preceding stage, the final VAT paid is made up of the sum of the VAT paid at each stage. Registered VAT traders are given a number and have to show the VAT charged to customers on invoices. In this way, the customer, if he is a registered trader, knows how much he can deduct in turn and the consumer knows how much tax he has paid on the final product. In this way the correct VAT is paid in stages and to a degree the system is self-policing.”

It is critical to note that the VAT mechanism fundamentally treats all entities in the same way - whether inside a particular jurisdiction or not. In the same way, a CAT would levy the same amount of tax on a good being produced inside or outside of a jurisdiction. This not only avoids double taxation, but counters any potential “leakage” effects, as there would be no carbon tax advantage to migrating production to another jurisdiction.¹³ So it also removes one of the major hurdles of carbon regulation – that is resistance from local industries.

This pressure were very clear in the latest round of European negotiations on setting emissions reduction goals. Specially in view of the economic environment, carbon-intensive industries, and manufacturing-oriented Eastern European countries lobbied against setting further restrictions on emissions. ¹⁴ As a result, European negotiators were forced to strike a compromise and allot various exemptions to various industries – arguably the most carbon intensive industries.¹⁵ In contrast, under a CAT regime carbon intensive industries would have little ground to protest. After all, a CAT would tax their competitors in foreign countries in the same way, and so it would not cause them to have a competitive disadvantage. And further more, the ultimate payers of the tax would be consumers of their products, rather than the producers themselves. So, far from being impractical, a CAT can even be more politically palatable than producer based carbon regulations.

The spread of the VAT through out the world also shows that the model is replicable, and that it can spread quickly:

“The rise of the value-added tax (VAT) is an unparalleled tax phenomenon. The history of taxation reveals no other tax that has swept the world in some thirty years, from theory to practice, and has carried along with it academics who were once dismissive and countries which once rejected it. It is no longer a tax associated solely with the European Community (EC). Every continent now uses the VAT, and each year sees new countries introducing it.”¹⁶

As such, given the history of VAT, there is little reason to believe that administering a CAT would be overly burdensome. ¹⁷

Accounting View: Counting Carbon for CATs

The second, and more important objection to a CAT relates to the difficulty of calculating the amount of embodied carbon in a product at any one point in the supply chain. Here the contrast with a VAT is indeed quite stark. A VAT is assessed on the actual value of a good purchased by a company - a value that is well known by both buyer and seller, and is already disclosed in contractual, and regulatory documents for tax and accounting purposes. In contrast, the embodied carbon in products is not only often unknown to both buyer and seller, but it can be very expensive to calculate precisely.

Specifically, the Life Cycle Analysis (LCA)¹⁸ methodology for calculating the total environmental impact of products is not only complex, but lacks a standardized methodology that can produce comparable results. Over the past few years, more and more companies have conducted LCAs to understand the product footprint of their products. But their experiences have been mixed. For example, a representative of Boots, a UK pharmacy which calculated the product footprint of some its products, later claimed that “there’s no way we can go through that for all of our products. The amount of work that would be required would be beyond our understanding and comprehension.”¹⁹

To simplify this expensive process of calculating embodied carbon, and thus make the CAT practical, three important challenges need to be overcome. First, a standard methodology needs to emerge. Second, relevant standard reference data tables must be created. And third, the CAT needs to incentivize a self-adjustment mechanism to further refine the inaccuracies in the methodology and reference data sets.

As I shall argue below, all three of these challenges can be tackled effectively.

1. Standard Methodology

The Carbon Trust, funded by the UK government, together with BSI (the British Standards Institute) and DEFRA (the UK Department for Environment, Food and Rural Affairs) have taken the lead in creating a standardized methodology for measuring embodied carbon. Together, they have developed…

“a Publically Available Specification (PAS) for the measurement of the embodied greenhouse gas emissions (GHG) of products and services across their life cycle. This is in response to market interest in understanding the contribution products and services make to climate change and a requirement for a standardised, consistent method organisations can use for this measurement… The PAS aims to provide a simplified approach that could be used by a greater diversity of product types and organisations. Key criteria for the method defined in the PAS is to enable its use in a range of applications to include for comparison and communication e.g. labelling, Green Supply Chain Management (SCM), Ecodesign etc.”²⁰

The resulting specification, called PAS2050, was published in October 2008²¹ following an extensive outreach program involving individual experts and institutions through out the world, and it lays out a basic framework for calculating the carbon footprint (ie the embodied carbon) in products. In effect, PAS2050 has laid the foundation for a GAAP of embodied carbon, allowing companies to use a standardized set of rules and procedures. And even if PAS2050 does not become the standard adopted through out the world, it can most certainly serve as a blueprint for other efforts, creating a path towards accepted standard methodologies for calculating embodied emissions.

2. Standard Data

All LCA analyses, including Carbon Trust’s simplified version, rely on “secondary data” – for example, reference data sets to estimate the environmental impact of either specific commodities or parts of each product – as opposed to “primary data”, which represents the direct measurement of emissions. For example, when measuring the embodied carbon of a computer, you could use “primary data” in calculating the “activity” related to that computer – say 50 Kilowatt hours of electricity used in the final assembly. You would then get primary data from your utility about the emissions related to each watt of electricity they deliver, to calculate the total emissions of the computer. If your local utility has not done a proper analysis of its emissions, you might use “secondary data”, such as national or regional averages of emissions per watt of electricity.

Similar processes are used to estimate the footprint of subassembly components. For example, if you are buying hard disk drives in the open market and have difficulties calculating the embodied carbon of such a drive, you could use a standard data sets, or other LCA studies of hard drives as proxies.

There are a number of data sets available in the market, and various praiseworthy efforts are underway to create more comprehensive data sets²², notably the European EXIOPOL effort²³. Unfortunately, most of these databases will inevitably remain somewhat limited, by industry or geography. There is no one reference data set for “secondary data” and even though PAS2050 does try to lay some rules to improve consistency, choosing different datasets can lead to drastically different results.

Despite these inaccuracies, the question we must really address is: Are these secondary data sets, in conjunction with standard methodologies like PAS2050 good enough for the assessment of a CAT? To a large extent, that depends on how a CAT is formulated and how it can incentivize stakeholders to refine their calculations over time.

3. Incentives / Self Adjusting Mechanism

A successful CAT would need to build a mechanism to incentivize companies to use as much primary data as possible, by imposing a penalty on the use of secondary data and datasets.²⁴ If the penalty for using a standard dataset compensates for any inherent inaccuracies in the reference dataset, then firms would be incentivized to measure embodied carbon with more and more precision, so as to avoid such penalties. In this way, any inaccuracies of the standard datasets would be neutralized by the proper incentive scheme.

Such a penalty would produce the right incentives in many ways. First, industries producing the most carbon intensive products would be incentivized to measure their footprints more accurately using primary data sources. If say, the tax for using a standard dataset (ie secondary data) is set to twice the tax for conducting a full analysis with primary data, carbon intensive products would be penalized far more than non-carbon-intensive ones in absolute terms. (That’s because, with this penalty, the higher the embodied emission of a product, the higher the absolute penalty for using secondary data.) Such a simple logic simulates a market mechanism means that the most carbon intensive industries would have the greatest motivations to measure their emissions more accurately using primary data. They would also be the most motivated to change processes and supply chains to reduce the carbon footprint of their products, and thus reduce their tax.

Second, the penalty creates a self adjustment mechanism to decrease the use of inaccurate data. If secondary data sets for a particular industry over-estimate the actual emissions of products in that industry, firms in that industry would be the most incentivized to start measuring and using primary data. Over time, as an increasing percentage of embodied product footprints become measured directly, penalties for using standard datasets can be increased, thus incentivizing a whole new set of industries to move towards primary data.

Third, within each industry, the penalties incentivize those companies which are producing their goods with fewer emissions than the average “standard”, to move towards direct measurement. The most carbon efficient firms in those industries would naturally push for more comprehensive and accurate methodological nuances using primary data, and they could use their lower footprints to differentiate their products.

One negative aspect of these penalties must be noted: They would obviously disadvantage small firms that don’t have the resources to conduct a full analysis of their products. However, as long as those small firms are not in carbon intensive industries, their CAT should be relatively low despite any penalties, even if they opt to just pay the penalty on using secondary datasets, and avoid making difficult calculations altogether.

Finally, there is one important argument for why the above incentives as administratively complex as they may seem, may be worth implementing. Perhaps the “problem” of calculating embodied emissions is not really the problem at all, but only a symptom of the greater problem at hand. If we do believe that carbon emissions risk harming our world significantly, then to tackle the problem, our first goal should be to understand and measure the sources of such emissions. How can it be resolved otherwise? So, measuring embodied emissions, far from being a “problem”, should really be the first step of the solution to the real problem: the fact that it is currently most difficult to account for carbon emissions world wide.

A Philosophical Note: Archimedes’ Lever

Archimedes²⁵, the Greek mathematician first known to describe a lever, famously said: “Give me a place to stand and I shall move the earth.” As stated at the beginning of this paper, one of the problems with inter-governmental treaties like Kyoto is that negotiators have little common ground to stand on - as they debate the merits of measuring production related emissions on a per capita or per GDP basis. But a CAT has the potential to create a lever for those concerned with carbon emissions, and empower them to affect production processes across the supply chain and through out the world.

The premise of this argument is that environmentally-aware consumers are disproportionally located in more developed nations that also represent an important and influential consumer base. So if standards related to CAT (for measuring embodied emissions) are implemented in European countries for example, they can impact products and production practices through out the supply chain.²⁶ Stiglitz, again, makes a similar argument²⁷:

“It would be easy to incorporate countries that failed to go along with the international regime. Producers in those countries would not be able to show carbon tax receipts. We could imagine two alternative regimes. One would follow the procedure just described: a tax would be imposed on the input on the assumption that it was produced in the most carbon intensive way possible. This by itself would provide a strong incentive for the country to impose a carbon tax at least on exports. The cost to outside buyers would be the same, but the producing country government would garner the revenue.

“Since most firms are unlikely to have two production lines—one for exports, one for domestic consumption—the tax would provide an incentive for reducing carbon emissions.”

The leverage afforded by CAT liberates us from the tyranny of inter governmental negotiations. In effect, it provides an alternative for countries or states that want to take action across the global supply chain without being at the mercy of other governments. If just one large US state like California imposes a CAT, it could have ripple effects through out production processes across the globe. That is because a CAT uses the global nature of the global economy in its favor and largely overcomes the issue of coordinating harmonized tax regimes across the world.

At the same time, international agreements will only enhance the CAT. If a country with a manufacturing base imposes a CAT, it exports would not be penalized at all, because of the way in which the CATs can be offset through out the supply chain. As such, developing nations will be incentivized to impose a CAT, even while developed nations need not depend on such adoption to get their own CATs implemented.

A Time to CAT?

It would have been unthinkable to propose a tax on carbon a decade ago, say at the time of the Kyoto negotiations. The public would have simply rejected any such notion. But a number of changes in the political environment have made the idea of a carbon tax much more acceptable today. First, it has become increasingly accepted among the scientific community and the public that carbon emissions cause global warming. Second, many notable economists and environmentalists have proclaimed that a carbon tax is the most efficient way to reduce carbon emissions.²⁸ (Paradoxically, China has become one of the first countries to impose a “green tax” on the import gas guzzling cars, albeit, possibly with the wrong motives.²⁹) Third, the European Emissions Reductions Plan is to migrate away from the current Cap and Trade towards a uniform price on carbon paid by all emitters after 2012 – effectively creating a carbon tax on local production thereafter, and paving the way for similar taxes on embodied carbon.

However, to-date, any discussion related to taxing carbon emissions outside of a particular jurisdiction has been limited to Border Tax Adjustments (BTAs). For example, BTAs have been mentioned as a possible addition to the post 2012 phase of the ETS. The European Emissions Reductions Plan announced in February 2008 suggested that “then big emitters inside the EU may be given their allowances free, or firms outside the EU could be forced to buy permits in the ETS, to apply to their imports.”³⁰ BTAs have also been referred to in Cap and Trade discussions in the United States. For example, America's Climate Security Act (ACSA), proposed by Senators Warner and Lieberman had a “clause that would penalise imports from countries that do not have an emissions cap. To get such goods through customs, importers would have to buy permits to cover the greenhouse gases emitted during their manufacture.”³¹ Perhaps as a result of this product discrimination, many such as The Economist have reacted to a BTA as “a dangerous weapon in the hands of America's growing gang of protectionists” to create trade barriers³². Even entities like the Carbon Trust have argued against BTAs because “almost any unilateral approach would be viewed with extreme suspicion by Europe’s trade partners. [And so] any attempt to impose border adjustments unilaterally would almost certainly be challenged.”³³

But yet again drawing a historical parallel to the VAT is most instructive, as the same kinds of fears were raised when the VAT was introduced in Europe. Indeed, according to a study by Ben Lockwood and John Whalley, when the VAT was first implemented in Europe, many feared that it would be viewed as a violation of existing trade agreements. “There was growing pressure on the US government to initiate a dispute settlement case on this issue… What transpired, however, was that no dispute settlement case was initiated and ultimately, no GATT negotiation took place on this issue... This was in large part because of contributions from the academic community which stressed that, in a very simple world where all consumption goods are taxed at the same rate and there is no savings and no labor-leisure choice, a movement between an origin and a destination basis in a VAT or sales tax will have no real effects on trade, production or consumption.”³⁴ The paper argues that similar arguments could be made in favor of BTAs, making them palatable from a trade perspective.

But current proposals for BTAs have targeted specific carbon intensive industries, and thus remain open to lobbying and political favoritism. On the other hand, the CAT

treats all companies in the same way. Not only does it not discriminate against foreign emitters, it is much more similar to the VAT in its uniformity and transparency. So, like the VAT, it is inherently fairer from a free-trade perspective.

Conclusions³⁵

This paper lays out the importance of embodied emissions both from a quantitative perspective, and from the perspective of moral equivalency. Simply put, if we care about carbon emissions at all, then the numbers show that we should also care about embodied emissions. I have also argued that tackling global emissions from a consumer responsibility perspective can help overcome some of the structural problems inherent in the current producer-responsibility based regulatory initiatives. The consumer-responsibility perspective is just more coherently attuned to the global nature of the world economy, and so it is a better fit to tackle disparate sources of emissions through out the world. It can also give tremendous leverage to environmentally minded jurisdictions to affect change globally by acting locally, and independently of inter-governmental negotiations. A Carbon Added Tax provides a clean mechanism for implementing such a consumer oriented scheme. If a CAT is structured properly, with the right incentive schemes, one could expect an emerging GAAP of carbon to become increasingly reliable and cheaper as more stakeholders adopt it. As such, the CAT can provide an effective platform for regulating carbon emissions world wide.

---

End Notes