Carbon footprint of the knowledge sector: what's the future?

Introduction

Attention of the entire world was focused on Copenhagen in December 2009 where world leaders and policy makers met under the auspices of the United Nations to develop policies to curb global warming. After a lot of discussions and deliberations an agreement was reached that recognised the need to limit global temperatures rising no more than 2°C (3.6°F) by 2020. The accord, reached between the USA, China, India, Brazil and South Africa, contains no reference to a legally binding agreement (BBC News, 2009).

Researchers have identified a number of sectors – the power sector being the most prominent one – as the major contributors to the emission of greenhouse gases. Pundits – academics, researchers and decision makers – from around the globe are working hard to generate facts and figures relating to the carbon footprint of different industries. In today's knowledge society and knowledge economy, we depend on access to, and use of, knowledge for every activity and decision making. But do we have a data set showing the carbon footprint of the knowledge sector – the knowledge production and distribution system; and do we have any alternative?

A detailed search on several databases – Library and Information Science Abstracts (LISA), ISI Web of Knowledge and many online journal databases – did not reveal any systematic study or research that can provide detailed figures of carbon footprint of the knowledge sector. In other words, finding hard data on the carbon footprint of the knowledge industry that can be used to conduct a systematic research to find ways and means for reducing the carbon footprint of the knowledge sector is rather difficult. This paper aims to provide some figures relating to the carbon footprint of the industries and activities that are related to knowledge creation, distribution and access. It looks at the carbon footprint of the traditional knowledge creation and distribution model, i.e. the carbon footprint of the printed book and journal industry, and also the carbon footprint of some activities associate with the traditional knowledge access and distribution system such as the photocopying activities undertaken within the provision of statutory licenses issues by national licensing agencies like Copyright Agency Limited, Australia (CLA), Copyright Agency Limited, UK (CAL), etc. Carbon footprint and environmental impact related to the creation distribution of digital content are also discussed.

Based on the CO2 emission figures presented in this paper, it is argued that a digital content service model would facilitate real time knowledge distribution and access at a much reduced environmental cost. Finally the paper proposes the need for further research towards developing a sustainable business model for digital knowledge transactions, and for studying user behaviour with regard to access and use of knowledge through a digital content service especially for knowledge‐intensive activities like education and research.

The knowledge industry

Although it is not highlighted and used in major public discussions, the size of the knowledge sector and its impact on the economy is quite significant. The following figures provide a glimpse of the size of the market:

• •

  In 2007 the value of the global book market was estimated to be US$127.5 billion which is estimated to rise to US$160.7 billion in 2012 (up 26 per cent from 2007); and in 2008, the US book publishers' net revenues reached US$40.32 billion (Healey, 2009)

• •

  In the UK the publishing industry, which is the second largest in Europe, has a turnover of over £18.4 billion, with its 8000 plus companies employing around 164,000 people and contributing to over 8 per cent of GDP (BERR, 2009).

• •

  The value‐added value of the copyright industry in Australia is approximately 100 billion Australian dollars which is about 10 per cent of the gross domestic product (GDP), and the annual compound growth of the industry is 4.7 per cent compared with 3.6 per cent of GDP (Price Waterhouse Coopers, 2008).

• •

  The creative industries in the UK produce in excess of ?70 billion of gross value added (GVA), the difference between output and intermediate consumption for any given sector/industry).

• •

  Its rate of growth is twice the average for the economy as a whole and it supports over 1.9 million jobs.

Knowledge distribution and access

Over nearly 500 years, since the invention of the modern printing press, the process of knowledge distribution and access has followed a typical business model where publishers have played a bridging role between knowledge creators and knowledge users. People have used one of the two (or both) ways of accessing knowledge – either by purchasing a knowledge product (a book, an article, a map, an illustration, research data and reports, etc.) or by using a social gateway to knowledge, namely, libraries. Thus knowledge products can be owned by a given user through purchase for example, or it may be owned by a library or an institution, where it can be shared by several users.

With the introduction and increasing use of computers in information handling several new methods and approaches have been developed for access to knowledge over the past five decades or so. Although the various knowledge products and their distribution and access mechanisms varied, the fundamental process remained more or less the same where a knowledge creator had to go to a knowledge producer (a publisher) who would create the knowledge product and would use a number of distribution mechanisms in order to make the product available to the users. However, this model began to change significantly with the advent of the internet where the barrier between the knowledge creator and knowledge user was broken resulting in a new paradigm where virtually every knowledge creator could be a knowledge producer (publisher) as well, and the need for formal distribution channels also became virtually non‐existent because users could get access to the knowledge product directly through the web. Of course this does not mean in any way that the role and importance of the publishers have become redundant, nor does it mean that every knowledge creator in today's world has become their own publisher. In fact, the web has brought new opportunities and challenges for the entire knowledge sector and it has influenced every player in the chain from knowledge creator to knowledge user.

The choice for a knowledge user in today's world is enormous – printed and digital knowledge products can be accessed through a myriad of means and platforms: through the bookstores (conventional and online), through online subscription, through a library, through various online e‐books, e‐journals databases, digital libraries, and off late through search engine services like Google Book. A knowledge product in today's world can be accessed and used for a price or a fee; it can be accessed for free under the Creative Commons licence; or it can be accessed and reused through statutory licences that are controlled by collecting societies like the CLA in the UK, CAL in Australia; and so on. Open archives movement, institutional repositories, Creative Commons licenses, free digital libraries and subject gateways, etc. have appeared over the past one decade or so with the conviction that knowledge should be available to everyone for free at the point of use. While the economic sustainability of these models, and even the prevailing publishing industry model is an issue, it is a fact that the web has opened up enormous opportunities for everyone in the knowledge sector, the challenge being that how best it can be adopted and adapted to make the knowledge creation and distribution process economically and environmentally sustainable.

The rest of the paper discusses the carbon footprint of the knowledge industry by pointing out to some data related to the carbon footprint of the publishing industry on the one hand, and some related practices in the knowledge‐intensive activities. Since detailed statistics relating to all the different activities in the knowledge sector are not available, this paper is built on the following sets of data:

• •

  Carbon footprint of a book, which provides an idea of the total CO2 emissions from the book industry that can be extrapolated to estimate the carbon footprint of the paper‐based knowledge products.

• •

  Carbon footprint generated from photocopying activities that are legally permitted under the statutory licences provided by licensing agencies like CLA, CAL, etc.

• •

  Carbon footprint for digital distribution of knowledge through the web.

• •

  Carbon footprint of digital access to knowledge.

Carbon footprint of the knowledge industry

Studies on the environmental impact of the content industry are few and far between. However, the data gathered from the available studies reports, presented below, pose some real causes for concern.

The BookScan UK Total Consumer Market trends for 2009 indicate that 235.6 million volumes of books were sold in the UK in 2009 with a total market value of £1.75 billion (Nielsen Book, 2010). Research Information (2009) data show that “24 trees are required to make one tonne of paper. In addition, 1.5 tonnes of coal are required to produce the electricity for one tonne of paper and this is responsible for emitting around five tonnes of carbon dioxide”.

Figures for the carbon footprints of books should consider CO2 emissions for the entire life cycle of books from production to distribution, storage, use and maintenance as well as disposal. According to one study an average book has a carbon footprint of 7.46 kilograms of CO2 over its lifetime (Kozak, 2003). A report from the Cleantech Group (Ritch, 2009, p. 3) indicates that:

UK‐based product design consulting firm Industrial Design Consultancy (IDC) used the LCA [Life Cylce Analysis] calculation in 2008 to estimate 8.16 kg of CO2 equivalent results from a children's book.

The Cleantech Study also reports two other studies, one of which estimates the CO2 emission from a book to be 10.2 kg while the other (rather conservative) estimation is 4.01 kg per book. Thus while the estimates vary, for the purpose of this paper the figure of 7.46 kg CO2 emission per book has been used which is also used by the Cleantech Group in their study of the environmental impact of Amazon Kindle (Ritch, 2009).

Taking the BookScan UK figure, 235.6 million books sold in the UK market would produce about 1.8 million tonnes of CO2. CARMA: Carbon monitoring for Action (CARMA: Carbon Monitoring for Action, 2009b) data show that the total number of power plants in Britain is 1,785 and together they produce 227 million tonnes of CO2 per year. So, CO2 emission from one power plant in the UK is 0.129 million tonnes. Thus the CO2 emissions for the books sold in the UK alone in a year will be equivalent to CO2 emission produced by nearly 14 power stations. The average carbon footprint per person in the UK is 10 tonnes (Climate Change Wales. n.d.). So, the annual CO2 emission for the books sold in the UK market is equivalent to the carbon footprint of 180,000 people.

A Bowker report estimated that 560,626 titles of books were to be published in the US in 2008 (Bowker, 2009). The book publishing industry report (Eco‐Libris, 2007) shows that 4.15 billion books were published in the USA in 2006. Another report (Poets and Writers, 2009) found the industry's annual carbon footprint is 12.4 million metric tons (11.27 million tonnes). This figure includes CO2 created by paper production and printing, transportation of books, landfill disposal of returned titles, and paper and energy use by retailers and publishers, combined with carbon storage lost when trees are harvested.

It is estimated that more than 1 billion books were printed in Germany in 2008 (Space Daily, 2009). So, at 7.46 kg, 1 billion books will produce 7.46 million tonnes of CO2. Therefore, the CO2 emissions from books published in Germany will be equivalent to the CO2 emissions from at least 62 British power stations. Carbon footprint of the book industry in the USA is several times bigger than that of the UK: about 11.27 million tonnes as opposed to 1.8 million tonnes. One way to understand the environmental damage caused by the publishing industry is to note that more than 30 million trees are cut down annually to produce the books sold in the USA alone (Eco‐Libris, 2007). CO2 emissions from the US book industry is equivalent to the emissions from 87.3 UK power stations. Thus, according to these calculations the combined annual CO2 emissions from the book market in the UK, USA and Germany alone is equivalent to the CO2 emissions from 163 UK power stations! Or alternatively CO2 emissions for book markets in the UK, USA and Germany is almost half of the annual CO2 emissions from all the power plants in Switzerland or Sweden (IEA statistics, p. 56); or it is twice as much of the CO2 emissions from all the power plants in Sri Lanka or the Sudan (IEA Statistics, 2009, p. 56).

While the total carbon footprint figures for other publishing activities, for example figures for the entire journal publication industry, are not available readily, figures from a study commissioned by a leading journal publisher Reed Elsevier give us some idea. The study noted that the total carbon footprint for producing the journal Fuel in 2007 was “just over 40 tonnes of carbon dioxide, and an ecological footprint of almost 21 global hectares” (Gough, n.d.). The figure in terms of the total carbon footprint of printed journals will be staggeringly high if the number of all the print journals in the world are multiplied by this figure! For example, Ulrich's Periodical Directory (2010) lists over 300,000 periodicals. Thus using the CO2 emission figures from the Reed Elsevier study, the annual CO2 emission from journal publishing will be 12 million tonnes. In fact, considering that many journals have more annual issues compared to the journal Fuel, and that these journals are published and transported across the globe, the total figure for CO2 emission can be way over 12 million tonnes. To put this in perspective, this is equivalent to the annual emissions produced by 94 power plants in Britain. It is true that out of the listed 300,000 journals in Ulrich's Periodical Directory, a significant number will be e‐journals and therefore should be taken out of this calculation. While this is true, not every journal has the same size and frequency similar to the Reed Elsevier journal Fuel, and the overall emission figure will depend on many factors like the size and frequency of journals, nature of content, circulation figures, and so on. Overall, the CO2 emission from the journal industry is therefore quite high.

From the above calculations it is quite evident that the current print models for production and distribution of knowledge is not environment‐friendly: the combined CO2 emissions from books produced in UK, USA and Germany, and that from world journal productions is almost equal to the annual CO2 emissions from all the power plants in Switzerland which is 42.18 million tonnes per annum (IEA Statistics, 2009, p. 56).

How can we reduce this carbon footprint?

The discussions so far in this paper clearly reveal that the current print models for production and distribution of knowledge, and the associated photocopying activities under statutory licenses, are not environmentally sustainable. Consequently a better and more environment‐friendly way for knowledge creation and distribution has to be developed and adopted, and this is a responsibility of not only the publishers, but also of all the stakeholders in the knowledge supply chain – from content creators and publishers to consumers.

There is no doubt that given the speed, robustness and versatility of the web, digital production and transactions would be much faster and that such transactions would take place on a real‐time as opposed to the current provisions where one may have to wait for weeks or even months for the printed knowledge product to be produced and shipped from one end of the world to another. As a result, as discussed below, a variety of digital content services have appeared over the past few years. However, the question remains whether digital content services help us reduce the carbon footprint and thereby make the knowledge sector more environment‐friendly?

Although further studies have to be conducted to find precise figures of environmental savings for each category of knowledge products and related activities, it may be argued that if we move to digital content services and therefore produce less number of printed products then there will be less carbon footprint. Digital content services will also reduce the number of pages that are currently photocopied under the statutory licenses (for example, under the licenses obtained from CLA), and this will also help save the environment. However, there will be environmental benefits only when it can be justified that digital content services produce less CO2 than the print‐based services.

Digital content services

Some recent developments for example, the Google Books, Amazon Kindle and similar e‐book readers, specific publisher‐ and bookseller‐driven digital book services, print‐on‐demand machines, etc. have brought some significant changes in the knowledge distribution and access mechanisms with promises for faster and easier access to knowledge. Side by side, myriad digital libraries, open archives, institutional repositories and subject gateways have also come up over the past few years that facilitate real‐time digital transactions of knowledge products. Perhaps the most prominent, robust and real‐time knowledge product transaction system is Google Books, which has over the past couple of years grown enormously and at the same time has drawn much attention, discussions, praises and criticisms all over the world.

Google Books has brought two major revolutions in the way we search and access content:

1. 1.

   it has invested a vast amount of resources to provide a full text search facility of in‐print, out‐of‐print and out‐of‐copyright books, and the collection of digital content is increasing rapidly through a massive digitisation program; and

2. 2.

   it allows the user to get access to the searched and retrieved content either through a library or through commercial transactions from publishers or book sellers.

Environmental benefits from a digital knowledge transaction system

Leaving aside the merits and criticisms facing the Google Book initiative or any specific publisher/aggregator/bookseller‐driven digital content services, it is evident that within a digital environment with less physical production of knowledge products, there will be less paper, less ink, less production and transportation of physical knowledge products, and all these will eventually result in a reduction of carbon footprint. However, although digital knowledge products do not produce additional carbon footprints for every copy that is being produced, there is a distribution cost, and it is important to consider what is the environmental cost associated with the online access, downloading and reading. So, environmental savings from a digital content services, as opposed to print‐based services, can be true only if digital transactions of knowledge products actually need less energy and therefore produce less carbon footprint.

While distribution costs do not disappear from the equation when we go digital, they are almost nothing compared to their printed counterparts. In the absence of any other comparable research data, one may use the tariffs charged by the Amazon S3 service (Amazon Web Services, 2009). The published tariffs on the Amazon S3 service site tell us that the charges for delivery of content should not be more than 17 cents per GB, and the cost goes down to 10 cents per GB for users who take up more than 150 TB a month. Now the question is how many books do we get for a GB of digital content?

Although the digital size of a book depends on a number of factors, such as the number figures, charts, etc., and their quality etc., the size of a Kindle e‐book version of a typical 300 page book can be less that 500 KB. So, using this rough estimate, the cost of delivering a typical digital book will be about a cent. This is almost nothing compared to the distribution cost of a physical book. In addition there will be a significant amount of indirect savings in terms of time, logistics and management operations, etc.

The next question is what is the environmental cost of data transfer for digital knowledge products?

While there is a bit of controversy amongst researchers as to how this should be calculated, one thing is sure that internet is used for numerous activities and transactions, and therefore the cost of the electronic infrastructure is being amortised. Furthermore, increasingly countries, especially in the developed world, are investing in broadband networks which will be there irrespective of whether we use the network for digital content transfer or not. Nevertheless, it will be important to know whether there are any additional environmental costs for data transfer and how the cost changes with the volume of data.

Data transfer across the internet and increases in the power consumption by the internet do not increase in a linear fashion. In fact, switches (the hardware that power the internet) use only a few more watts at full load than they do when they are completely idle. For example, at idle the Cisco Catalyst 2960 and HP ProCurve 2650 draw 33 watts and 51 watts respectively, and yet at full load they draw 37 watts and 57 watts respectively (Cisco, 2009). Given that switches can handle gigabytes of data per second, full load increases power consumption only by a few watts, and because digital books tend to be very small (the size of a Kindle e‐book version of a typical 300 page book can be less that 500 KB), the increased energy usage, and the corresponding carbon footprint for digital content transfer seems negligible.

So whether or not people are using the internet for digital content transfer, the power is still being consumed, and with the digital transactions only a small amount of additional electricity is used. In effect, the environmental cost of data transfer through the internet is negligible as is the financial cost for digital delivery of knowledge products compared with the financial and environmental costs associated with the distribution of physical knowledge products.

The next question is how much CO2 is produced for online reading of digital content. In order to answer this question we need to know how much emission is produced by a typical online reading device like a laptop or a desktop computer.

Fossil fuelled electricity generation has the largest carbon footprint (up to 1,000 g CO2/kWh) and low carbon technologies have low life cycle carbon emissions (<100gCO2eq/kWh) (Parliamentary Office of Science and Technology, UK, 2006). Taking all kinds of power plants into consideration it may be estimated that on an average 500 g of CO2 is emitted for 1 kWh of electricity; this figure corroborates the findings of the Climate Change Wales website (Climate Change Wales, n.d.).

The average electricity consumed in a laptop is 32 watts/hr, and thus it emits 0.016 kg of CO2/hour. Since an average book produces 7.46 kg of CO2, on an average emission from a typical book will be equivalent to reading an online book for 466 hours on a laptop. An average desktop operates at 110 watts. Using similar calculations it may be noted that the carbon footprint of a typical book will be equivalent to reading the same book in digital form for nearly 150 hours on a desktop.

Thus digital knowledge products cost almost nothing in the distribution chain and they have a much smaller environmental footprint for online reading, and moreover digital versions will have more flexibility for real time transfer and access, value addition and use in a shared digital environment.

Conclusion

Based on the data of CO2 emissions of physical books and their digital counterparts, it is possible to get some comparative figures as follows:

• •

  A typical book (of about 300 pages) creates 7.46 kg CO2.

• •

  Reading a similar size book on a laptop for about 466 hours will produce the same amount of CO2. Since a general reader will need much less time to read a typical book, the savings can be quite significant.

• •

  Keeping in view the negligible cost for distribution of digital copies, and the very small environmental cost for reading books online, the environmental savings will be quite significant.

However, further research is needed in order to build and use such a digital knowledge transaction model. The most important aspect of this research should be to develop and test a sustainable business model. Side by side the business practices in the knowledge industry have to change from a product‐based to a service‐based model. An early version of such a digital content transaction model has been proposed recently in the literature (see for example, Chowdhury, 2009a, b). However, the model needs to be tested in order to generate comparative figures for its benefits both in terms of its economic and environmental sustainability. Furthermore, since this will bring a paradigm shift in the entire knowledge sector – creation of knowledge products and their distribution, access, use and super distribution – more research needs to be undertaken to study its impact on various stakeholders in the knowledge chain – on the knowledge creators and producers (publishers), on the knowledge consumers – users and downstream users (for example academics and students in a class who denote the two classes of users respectively), on the libraries and intermediaries like online databases and search service providers, on the legal frameworks such as the intellectual property and copyright laws, and most importantly on the society as a whole. A low cost – both financially and environmentally – easy‐to‐use, flexible and real time content service model that is built on the latest and appropriate web technologies and standards will lead the world to a true knowledge society.