Simulating the impact of new industries on the economy: The case of biorefining in Australia

Ecological Economics 107 (2014) 84–93

Contents lists available at ScienceDirect

Ecological Economics journal homepage: www.elsevier.com/locate/ecolecon

Analysis

Simulating the impact of new industries on the economy: The case of biorefining in Australia Arunima Malik a,⁎, Manfred Lenzen a, Rômulo Neves Ely a,b, Erik Dietzenbacher c a

ISA, School of Physics A28, The University of Sydney, NSW 2006, Australia Energy Planning Program, Graduate School of Engineering, Federal University of Rio de Janeiro, Centro de Tecnologia, Bloco C, Sala 211, Cidade Universitária, Ilha do Fundão, Rio de Janeiro, RJ 21941-972, Brazil c Faculty of Economics and Business, University of Groningen, P.O. Box 800, NL-9700 AV Groningen, The Netherlands b

a r t i c l e

i n f o

Article history: Received 16 September 2013 Received in revised form 24 July 2014 Accepted 31 July 2014 Available online xxxx Keywords: Input–output analysis Hybrid LCA Biofuel Ethanol Sugarcane Employment

a b s t r a c t We investigate the economic and employment consequences of introducing a new sugarcane-based biofuel industry into Australia. We model the new biofuel industry on the production recipe of the existing large-scale gasoalcohol and alcohol sectors in the Brazilian economy. To this end we utilise a hybrid IO-LCA (input–output life cycle assessment) approach, which involves inserting data on new processes and/or sectors into an existing IO table. In particular, we develop and test an analytical and a numerical approach for re-balancing an IO table augmented with rows and columns representing large new biofuel industries. We quantify changes in economic output and employment in the Australian economy. We conclude that a future biofuel industry will be employment-positive for Australia. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Decision-makers in Australia are increasingly concerned about energy security, and one of the scientific responses considered are domestic biofuels. At present, Australia is in the early days of a biofuel industry. To successfully enhance biofuel opportunities in Australia, a myriad of technological changes are needed. For example, present fleet of cars in Australia cannot run on ethanol fuel. Like Brazil, Australia will need to welcome flex-fuel vehicles that can run on both petrol and ethanol. The technology for the construction of these vehicles is currently not available in Australia. Thus, the introduction of a biofuel industry will require substantial technological change in the economy. The economic repercussions of technological change have been a topic of research for many years. Researchers often use input–output (IO) techniques to understand the consequences of introducing new products and/or industries into an economy.1 Rose (1984) provided an overview of the different methods for estimating technological change in IO matrices. One variant of these methods is the

⁎ Corresponding author. Tel.: +61 2 9351 5451. E-mail address: [email protected] (A. Malik). 1 In addition to IO techniques, econometric and CGE models are available as well. However, we chose the IO model because IO models adopt parameters that are directly measurable and that are able to reflect technologies in great sectoral detail.

http://dx.doi.org/10.1016/j.ecolecon.2014.07.022 0921-8009/© 2014 Elsevier B.V. All rights reserved.

augmentation approach, which involves inserting data on new processes and/or sectors into an existing IO table. For example, Lave et al. (1995) used this method to compare electricity use and toxic emissions of paper cup and plastic cup production. Joshi (1999) extended their approach to assess the environmental performance of steel and plastic automobile fuel tank systems. The augmentation approach is, undoubtedly, useful for modelling the effects of introducing new products and/or industries into an economy. However, this approach disturbs the IO balance of the IO table. If the size of the new sector is small, the disturbance is negligible and the augmentation causes a marginal change for the total output of sectors. For this reason, neither Joshi (1999) nor Acquaye et al. (2011) nor Wiedmann et al. (2011) re-balanced their modified IO table after augmentation. The IO system is significantly disturbed if the insert represents a large portion of the economy. Probably the most explicit treatment in the existing literature is Suh (2004), (Appendix A) who proposed a method for reconciling IO accounts and process data inserts within integrated hybrid life cycle assessment (Heijungs and Suh, 2002). Whilst Suh employed a manual procedure, Li et al. (2012) and Liu et al. (2012) more recently used a RAS approach. Stone (1961) developed RAS, also known as bi-proportional matrix balancing. It works by scaling the rows and columns of an initial matrix using prescribed row and column sums (marginal totals) to obtain an updated matrix (Bacharach, 1970; Miller and Blair, 2009). Since Stone's initial conception, many RAS variants have been developed and applied

A. Malik et al. / Ecological Economics 107 (2014) 84–93

in IO studies, for example, TRAS (Gilchrist and St. Louis, 2004, relaxing the restriction to marginal totals as the only constraints), GRAS (Junius and Oosterhaven, 2003, extending to negative elements), and KRAS (Lenzen et al., 2009, allowing non-unity coefficients and conflicting constraints). Li et al. (2012) utilised RAS to successfully rebalance a 2007 IO table for China, augmented by a large wind energy sector. Liu et al. (2012) added 11 sectors into the 2006 IO table of Taiwan, and used the KRAS approach for balancing. The EXIOPOL project (Tukker et al., 2013) followed a two-step procedure, where first domestic supply–use table blocks were balanced, which in turn were then trade-linked into a complete multi-regional input–output table (MRIO). RAS has also been used to study technological change in IO systems that were not augmented with new rows or columns, but only perturbed by updating individual cells. Van der Linden and Dietzenbacher (1995, 2000) argued that RAS can be used for measuring technological change, and developed a method for decomposing these changes in the IO matrix into row, column and cell-specific effects. Dietzenbacher and Hoekstra (2002) employed the RAS method for studying the effects of technological change and trade in the Netherlands over the years 1975–1985. Andreosso-O'Callaghan and Yue (2000) used RAS to investigate the industries responsible for causing structural change in the manufacturing sector of the Chinese economy. Dobrescu and Gaftea (2012) estimated the technical coefficients of the Romanian economy to test the applicability of the RAS procedure in an emergent system. Traditionally, both the augmentation approach and the cellupdating approach change the production recipe, which is the structure of a use matrix column, of at least some sectors. Whilst this effect cannot be avoided in the cell-updating approach, the augmentation approach can be modified to keep the production recipe of all existing sectors intact. The rationale for retaining production recipes is that, generally speaking, there is no reason why the introduction of a new sector into an economy should cause changes in the inputs of already existing sectors. In our case, introducing a new biofuel industry into the Australian economy does not require any industry to change the way they produce, except for purchasing biofuels instead of the previously used petrol. This is intuitively clear: Even after biofuels have entered the market, any power plant will use the same proportion of coal, any farmer the same proportion of fencing, and any school the same proportion of paper, in their inputs. This is necessarily a simplified (ceteris paribus) view of how an economy would adjust in real-life, for example because we have assumed that biofuels would be a perfect substitute for petrol, and would fetch the same price, i.e. would not require any subsidy.2 In reality, introducing a new sector would likely cause relative prices to change, and as a result sectors would adjust their production structures (see for example the work by Duchin and Lange, 1992, and Duchin and Levine, 2011). If such effects would be expected to be large, one would have to adopt a different type of model, such as a Computable General Equilibrium (CGE) model, or a linear-programming type model such as the WTM/RCOT model by Dilekli and Duchin (accepted for publication) in which the choice between several technologies is kept endogenous, and is solved via minimising total factor inputs. Our work has two main goals: Our first aim is to develop and test an analytical and a numerical approach that allow the re-balancing of an IO or supply–use table that was unbalanced by an augmentation with rows and columns representing large new industries and/or technologies. The novel characteristic of these methods is that they act only on columns and thus keep the production recipe of economic sectors constant. The numerical approach is a modified RAS method that serves to analyse the step-by-step repercussions of the initial 2 Using a linear programming type of IO model, Dilekli and Duchin (accepted for publication) study the introduction of a new biofuel sector in the US. Using the biofuel's column as given (instead of rebalancing) they obtain the subsidy endogenously.

85

insertion of new industries and/or technologies. Our second aim is to investigate the consequences for total output and employment of introducing a new sugarcane-based biofuel industry into Australia. The utilisation of both an analytical and a numerical, iterative approach is useful because they inform, respectively, about the final outcome of the introduction of a new biofuel industry, and about the economy's adjustment trajectory following the initial introduction. We model the new biofuel industry by taking the production recipe of the existing large-scale gasoalcohol and alcohol sectors in the Brazilian economy. In the following section we will first explain our approach in detail. In Section 3 we will apply our procedures to a case study of introducing a new sugarcane-based biofuel industry into the Australian economy. We discuss our findings in Section 4 and conclude in Section 5. 2. Methodology Our approach encompasses four steps: 1) we augment an existing supply–use matrix by a number of empty rows and columns; into these rows and columns we insert data for new industries and products, which will unbalance the table (Section 2.1); 2) we re-balance the table using an analytical adjustment approach; and analyse the difference between the state of the economy preand post-adjustment (Section 2.2); 3) we use a RAS-type numerical approach to obtain details on the trajectory of the economy between the pre- and post-adjustment stages (Section 2.3); and 4) we quantify the impacts of the new industries on the economy in terms of monetary output as well as employment (Section 2.4). We will now describe these four steps in detail. We will use the example of our case study described in Section 3. All calculations were carried out in the Industrial Ecology Virtual Laboratory (IELab) on the NeCTAR research cloud (Lenzen et al., 2014). The IELab is a unique research platform that offers high level of automation for analysing input–output data. 2.1. Augmentation of the Transactions Matrix In our case study (for further details see Sections 3.1 and 3.2) we appraise two new industries and products: a) alcohol made from sugarcane, and b) gasoalcohol made from a mix of sugarcane-based alcohol and conventional petrol. The gasoalcohol replaces part of the demand for conventional petrol. In a first step, we set the percentage π to which conventional petrol sales shall be reduced, and adjust 1

0

U petrol;i ¼ ð1−πÞU petrol;i ; ∀i≠alcohol; gasoalcohol 1 ypetrol

0

¼ ð1−πÞypetrol

;

ð1Þ

where U denotes elements of the use matrix, y denotes elements of the final demand matrix, and the superscripts 0 and 1 denote matrices pre- and post-augmentation. This step is represented by the light grey fields in the use (U) and final demand (y) matrices shown in Fig. 1. Similarly, we adjust the production of the conventional petrol sector: 1

0

V petrol;petrol ¼ ð1−πÞV petrol;petrol ;

ð2Þ

where V denotes elements of the supply matrix. This step is represented by the light grey field in the supply (V) matrix in Fig. 1.

86

A. Malik et al. / Ecological Economics 107 (2014) 84–93

In principle, this unbalanced matrix could be re-balanced using a RAS approach. However, the row scaling operations of a RAS procedure would change the production recipes of sectors beyond the mere substitution of gasoalcohol for petrol. As justified in the introduction, there is no reason why these production recipes should change just because a new sector exists in the economy. Therefore, we require some means of balancing the unbalanced table by just changing output3 levels (row and column sums) and sales patterns (row structures), and not production recipes (column structures). To this end we describe two approaches in the following subsections – one analytical and one numerical – for balancing the augmented supply–use system whilst retaining the production recipe of all sectors. 2.2. Analytical Approach Suppose – without loss of generality – that we have just two sectors, sugarcane (s) and petrol (p), and that the supply matrix V0 is diagonal. The input coefficients matrix (A) and final demands (y) are then given by " 0

Fig. 1. Schematic diagram showing the augmentation of an existing supply–use table with two new industries and two new products. U: use matrix; V: supply matrix; v: value added, net taxes and imports; y: domestic final demand and exports.

Then we let gasoalcohol fill the gap in petrol sales: 1

0

U gasoalcohol;i ¼ πU petrol;i ;

∀i≠alcohol; gasoalcohol

1 ygasoalcohol

¼

0 πypetrol

1

0

ð3Þ

ð4Þ

Following, we construct the input recipe of the gasoalcohol sector as a mix of alcohol and petrol (black fields in Fig. 1), according to 1

1

U alcohol;gasoalcohol ¼ αV gasoalcohol;gasoalcohol

ð5Þ

and 1

1

U petrol;gasoalcohol ¼ ð1−α ÞV gasoalcohol;gasoalcohol ;

ð6Þ

where α denotes the proportion of alcohol in gasoalcohol. The gasoalcohol sector is thus functioning as a simple re-router of certain (variable) proportions of alcohol and petrol into a compound product. Finally, we insert the intermediate inputs U1i,alcohol and primary inputs v1i,alcohol of the alcohol sector, taken from the Brazilian IO data, as well as the total supply 1

V alcohol;alcohol ¼

X i

1

U i;alcohol þ

X

1 v i i;alcohol

0

Asp

0

App

Aps

ð7Þ

into the dark grey fields in Fig. 1. Finally, we take U1alcohol,alcohol = U1gasoalcohol,alcohol = U1gasoalcohol,gasoalcohol = 0. 2.1.1. Balancing Issues The way in which the augmented supply–use table is constructed (Eqs. 1–7) means that – except for the petrol refining industry – inputs and outputs of industries are balanced, that is xij = zij ∀ j ≠ petrol. In contrast, demand and supply of products are not balanced – except for gasoalcohol, that is xpj ≠ zpj ∀ j ≠ gasoalcohol.

0

#0

0

0

ys 0 yp

0

and y ¼

! :

ð8Þ

with, for example, A0sp = U0sp/Vpp. In the augmented matrix we have two additional sectors, alcohol (a) and gasoalcohol (g). Given the parameters α and π (see Section 2.1 for explanation), augmentation of the coefficient matrix yields 2

;

(represented by the grey fields in the use (U) and final demand (y) matrices in Fig. 1), as well as the gap in the petrol production (represented by the grey field in the supply (V) matrix in Fig. 1): V gasoalcohol;gasoalcohol ¼ πV petrol;petrol :

A ¼

Ass

0

Ass 6 0 1 6 A ¼ 6 ð1−πÞAps 4 0 0 πAps

0

Asp 0

ð1−πÞApp 0 0 πApp

Asa Apa 0 0

3 0 0 0 ys 7 B 1−α 7 1 B ð1−πÞy0p 7 and y ¼ B @ 0 α 5 0 πyp 0

1 C C C: A ð9Þ

The total output of the augmented table is x1 = (I − A1)−1y1, after ^ 1 , where the hat which the balanced IO table is obtained from A1 x symbol ‘^’ denotes vector diagonalisation. Except for changes introduced by the insertion of the new sectors, the production recipe of all sectors remains unaltered. Lenzen and Rueda-Cantuche (2012) provide detailed methodology for obtaining the direct coefficient matrices and Leontief inverses from supply–use tables. The difference between the pre- and post-adjustment state of the economy can be described by 2(x1i − x 0i )/(x 1i + x 0i ), where x0 = (I − A0)− 1y0 is the total output of sector i before balancing. 2.3. Numerical Approach 1

^ can be achieved using an iterative, RASA result identical to A1 x type numerical approach. We are interested in this additional approach because we would like to investigate not only the final outcome of introducing gasoalcohol, but also the stepwise changes that the Australian economy would have to undergo in order to arrive at this outcome. ^ 1 with a supply–use In the following section we will associate A1 x structure. In a first step, we balance supply and demand of products. To this end we compute total demand of products p

N

M

x ¼ U1 þ y1 ;

ð10Þ

3 Because of the new alcohol sector some industries may have a change in their production. Due to scale effects, this may change their input proportions. The empirical results in this study indicate that changes in other sectors' production are relatively minor.

A. Malik et al. / Ecological Economics 107 (2014) 84–93

where N is the column dimension of the use matrix, M the column dit

N

M

mension of the final demand matrix, where 1 ¼ f1; 1; …; 1g and 1 ¼ |{z} N elements t f1; 1; …; 1g are N-sized and M-sized row summation operators, i.e. N|{z} M elements

and M-element column vectors with ones, and where the superscript t denotes transposition. We also compute total supply of products (p) according to p

Nt

z ¼ 1 V:

ð11Þ

87

den Bergh, 2006; Leontief and Ford, 1970). These quantities often include CO2 emissions, water, land and energy use, but also, as in our work, employment. Let Q be a satellite account containing the employment Qi of industry i. 1

Then, the vector q ¼ Qˆ x 0 describes the employment intensity qi of industry i as the employment per unit of total output. Following standard IO practice, proportionality is assumed between industry output and employment, the sectoral employment impacts of the introduction of the new biofuel sector are then:   ^ x1 −x0 ; ΔQ ¼ q

ð16Þ

In a second step we scale total supply of products according to V ij →V ij

xpj zpj

:

ð12Þ

After execution of this step, demand and supply of products is balanced, but inputs and outputs of industries are not balanced. The third step therefore involves industry balancing. To this end we now compute total inputs of industries (i) i

Nt

Kt

z ¼ 1 U þ 1 v;

ð13Þ

where K is the row dimension of the extended value-added block, and where 1K ¼ f1; 1; …; 1gt is a K-sized row summation operator. We |{z} K elements

also compute total industry output according to i

N

x ¼ V1 :

ð14Þ

Since the supply matrix was scaled up during the balancing operation in Eq. 11, we expect total industry output xi to exceed total industry input zi. In a fourth step we scale total industry input according to i

U ij →U ij

xj zij

i

and vij →vij

xj zij

:

ð15Þ

After execution of this step, inputs and outputs of industries are balanced, but demand and supply of products is not balanced, and the algorithm returns to Eq. (10). Note that throughout the procedure, scalers x/z are only ever a function of column indices j, and never a function of row indices i. In other words, we only ever scale entire columns, and never rows. This mathematical feature embodies the intention of our modification: to ensure that production recipes of sectors remain unaltered during table balancing. In the literature, the changes caused by row scalers are termed substitution effects, because they demonstrate the degree of substitution of one input for another over time (Miller and Blair, 2009). For example, if metal products were replaced by plastic over time, subsequently the elements in the row indicating plastic products would increase, whereas the elements in the row representing metal products would decrease (EEC, 2008). In essence, since in our study, we only scale the columns of an augmented IO matrix, we do not allow substitution effects during the rebalancing of the augmented matrix. Recall that substitution of petrol by gasoalcohol was explicitly modelled in Eqs. (1) and (3). Our rebalancing procedure thus only changes the sales structure of the sectors. As mentioned in the introduction, a new product entering the market, generally, does not change the inputs needed to produce other products (see also footnote 2). Therefore, our modified RAS-type numerical approach preserves the production structure of all sectors. 2.4. Measuring Employment Impacts

where once again the hat symbol ‘^’ denotes vector diagonalisation. As the gasoalcohol sector is functioning as a simple re-router of certain (variable) proportions of alcohol and petrol into a compound product, we have not allocated any employment to this sector. 3. Case Study: A Future Sugarcane-Based Alcohol Industry in Australia 3.1. Background Brazil is the leading sugarcane producer in the world. During the financial year 2011–12, approximately 557 million tonnes of sugarcane were produced. Almost half of this amount was used as feedstock to produce 22 billion litres of ethanol (Barros, 2012), which is regarded as a suitable alternative fuel to petrol. In 1997, the Brazilian government released mandatory guidelines to blend anhydrous ethanol with petrol (Schmitz et al., 2011). Since then, no vehicle is allowed to run on pure petrol. Gasoalcohol, a blend of 20–25% anhydrous ethanol4 and 75–80% pure petrol, is presently used for running petrol engines (Macedo et al., 2008). Brazil became the first country to introduce flex-fuel vehicles in 2003, which can run on either hydrous ethanol or any proportion of anhydrous ethanol and petrol blends (Macedo et al., 2008). Since their induction, the fleet of flex-fuel cars has grown to more than 16.3 million and the sales constitute approximately 90% of total vehicle sales in Brazil (Barros, 2012). The flex-fuel technology proved highly successful, as the users have freedom of choosing either pure ethanol or its blends for running their vehicles. The success of this technology received worldwide attention. In Australia, General Motors trialled Saab 9-5 Biopower E855 flex-fuel vehicles in 2007 (Green Car Congress, 2013). However, the interest in these vehicles was largely subdued due to the widespread unavailability of ethanol fuel in Australia. Australia produces about 35 million tonnes of sugarcane per year, the majority of which is used for the production of sugar (Cane Growers, 2010). Nearly 80% of manufactured raw sugar is exported, making Australia the third largest raw sugar supplier in the world (DAFF, 2012). In 2012, about 14%6 of the Australian sugarcane was used as feedstock to produce less than 400 million litres of ethanol (ACCC, 2013). In contrast, nearly 20 billion litres of petrol was consumed by motor vehicles in 2012 (ABS, 2013), causing 50 million tonnes of carbon dioxide equivalent (Mt CO2-e) emissions. To reduce carbon emissions caused by petrol combustion, countries worldwide, such as Brazil, are shifting towards renewable fuel sources. 3.2. Case Study Description The success of the Brazilian production recipe for producing sugarcane-based ethanol is well known. The question arises whether this recipe could be feasibly implemented in the Australian economy. The aim

4

Hydrous ethanol is 100% alcohol, with no petrol content. E85 is a blend of 85% denatured ethanol and 15% petrol. 6 The percentage was calculated using the conversion: 1 tonne of sugarcane yields about 85 l of ethanol (SugarCane.org, 2013). 5

The economic IO framework can be extended to incorporate physical quantities (Duchin, 2004; Giljum and Hubacek, 2004; Hoekstra and van

88

A. Malik et al. / Ecological Economics 107 (2014) 84–93

of our case study is to evaluate the consequences of introducing sugarcane-based Brazilian alcohol and gasoalcohol manufacturing into the Australian economy. To this end, we insert the production recipe of alcohol into an Australian supply–use table assuming that gasoalcohol will substitute the demand for petrol. Gasoalcohol consists of between 20% and 25% of anhydrous ethanol7 (Coordenação-Geral de Açúcar e Álcool, 2011). We therefore set α = 0.25. In our case study we investigate a scenario where sugarcane-based gasoalcohol would replace π = 20% of current (2009) demand for conventional petrol, or 2.3 billion litres (AIP, 2013). At the price of 1.13$/l, 2.3 billion litres equate to 2.6 billion dollars. Therefore, we reduce the sales of conventional petrol by 20% (from 10$bn to 8$bn), and let gasoalcohol fill the gap. Then, we construct the input recipe of gasoalcohol as a mix of 25% alcohol and remaining petrol (see Section 2.1 for stepwise explanation of the augmentation process). We balance the table using an analytical approach to preserve the structure of the sectors' production recipe (see Section 2.2), and run our RAS routine (Section 2.3) to understand the sequential effects of supply-chain succession of changes in employment and total output. 3.3. Data Sources and Preparation For populating matrices U, V, v and y in Fig. 1 we use a balanced Australian supply–use table for the year 2009, taken from the Eora MRIO database (Lenzen et al., 2013), based on data in ABS (2012a, b), and further described in (Lenzen et al., 2012, 2013) . We compress competitive imports, taxes on products and subsidies on products into one row each, and append these rows to the conventional value-added block. The dimensions of this supply–use system are N = 345, M = 7, and K = 11 (Fig. 2). The production recipe of a new Australian alcohol sector (intermediate inputs U1i,alcohol and primary inputs v1i,alcohol) was estimated by a) taking the production recipe of the existing (large-scale and mature) Brazilian alcohol sector (IBGE, 2011) in purchaser's price, and converting the data to basic prices (Guilhoto and Sesso Filho, 2005, 2010), b) converting the currency from Brazilian real to Australian dollars using the average exchange rate for year 2009 (1 BRL = 0.64 AUD) (X-RATES, 2009), c) constructing a concordance matrix C for bridging the Brazilian and Australian sector classifications, and finally d) converting the Brazilian production recipe according to "

1

#

Ualcohol 1 valcohol

" ¼C

Brazil

Ua0 lcool Brazil

va0 lcool

# :

ð17Þ

After augmentation the supply–use system measures N = 347. 4. Results 4.1. Changes in Sectoral Output Adding Brazil's alcohol production recipe into the Australian IO table created an imbalance in the Australian IO table. We re-balanced the augmented table using the analytical approach (Section 2.2), and calculated the change in total output of selected sectors. The results are given in Table 1. Taken together, the two new sectors (gasoalcohol and alcohol) produce 3.2$bn. As expected, the sugarcane sector shows a substantial increase in output, given it is the prime input into alcohol refining, representing some 40% of its inputs. The petrol and diesel sector reports the greatest decrease, which is an aftermath of 20% reduction in the sales of petroleum. This decrease is just in excess of 5%, which can be explained by a combination of a π = −20% reduction in pure petrol 7 Which we refer to as “alcohol”, or the Portuguese “álcool” as listed in the Brazilian supply–use tables.

production, and a (1 – α)π = 3/4 × 20% = + 15% contribution to the gasoalcohol mix. There is a positive change in the output of industries providing direct inputs into the alcohol refinery. The opposite is true for the direct suppliers of petrol. Table 1 presents the results for a selection of the 347 sectors. In the following we explain the contributions of these top sectors. Ethanol is produced using sugarcane as well as by-products (molasses and sugarcane bagasse) of processing raw sugar into refined sugar (Olbrich, 1963). These by-products are vacuum-pressed into canvas bags before transporting them to an alcohol refinery (Zhu et al., 2011). The agricultural services sector provides services for growing and harvesting sugarcane, for example the renting-out of large harvesters used for chopping and accumulating the cane into parcels vehicle ready for road transport to the alcohol refinery. Before processing, the quality of cane is analysed in the laboratory, and the spectrophotometric and chromatographic data are provided by data processing businesses (Tewari and Malik, 2007). The sugarcane is then transferred to a crushing system via conveyer belts. The crushed cane is subjected to rollers or diffusion systems to separate bagasse from cane juice. The repairs sector deals with the repairing of small hand tools such as drills, and different components of industrial machines, for example mechanical seals on pumps. During the industrial process, sugarcane bagasse is first hydrolysed and converted to sugars (Sun and Cheng, 2002). Afterwards, bagasse, cane juice and molasses undergo fermentation and distillation processes to produce ethanol. The fabricated metal products sector supplies locks, fittings, nuts, bolts and other small hand tools to the refinery. Instruments such as distillation columns, condensers and fermentation tanks need regular cleaning to remove build-up of impurities. Acid cleaners, sanitisers and disinfectants sourced from the chemical products industry are used for this purpose. The paper products sector supplies paper for office and sanitary use. Personnel in the refinery follow strict ethanol production protocols. One of these protocols includes guidelines for converting hydrous ethanol to anhydrous ethanol. A process called chemical dehydration removes water from hydrous ethanol using quicklime (Kumar et al., 2010), which is supplied by the non-metallic mineral products sector. The alcohol refinery obtains business services, such as research-and-development consulting, and education and training of staff. Contrary to the alcohol supply chain, petrol refining experiences a decrease in output. Petrol is produced by fractional distillation of crude oil, which therefore constitutes 74% of the direct requirements of the petrol and diesel sector. As a consequence, a 5% decrease in the demand for petrol results in an almost 2% reduction in the output of the crude oil sector. The petroleum and coal products sector supplies solvents, which are used to separate refinery gas into its individual components (NZIC, 2008). A chemical process called alkylation is used to convert light hydrocarbons into heavier ones using chemicals, such as sulphuric acid and hydrofluoric acid. The heavier product consists of isooctane, which has excellent anti-knock properties in petrol engines (Islam et al., 2012). Australia imports crude oil from Middle Eastern countries (ACIL Tasman, 2009), which is transported via ships in large storage containers. After the oil is imported into Australia, it goes through the hands of wholesalers before it is made available to the refineries. 4.2. Changes in Employment The total (direct & indirect) employment effects, of introducing a sugarcane-based alcohol sector into the Australian IO table, can be examined by following the methodology outlined in Section 2.4. Fig. 3 shows the top and bottom 8 sectors as a function of net change in employment (measured in full-time equivalents, FTE) resulting from the 20% substitution of petrol, with gasoalcohol (see Eq. (16)). These sectors are aggregated over groups of similar sectors. The

A. Malik et al. / Ecological Economics 107 (2014) 84–93

89

Fig. 2. Heat map of the augmented supply–use system used in this study. x- and y-axes show sector numbers. The entire system measures (347 + 347 + 11) × (347 + 347 + 7) = 705 × 701 sectors. Grey shades represent the log10 of transaction values expressed in '000 Australian Dollars. The supply–use structure adheres to the schematic representation in Fig. 1. U: use matrix; V: supply matrix; v: value added, net taxes and imports; y: domestic final demand and exports.

results for a larger set of sectors are given in the last column of appendix Table A1. The sectors recording an increase in output (Section 4.1) also experience job gains, and vice versa (which is because of the proportionality between sector output and employment). The alcohol sector requires personnel to set-up and operate the refinery. As the area of plantation increases, additional farmers, harvester operators and tractor drivers are needed in the sugarcane and agriculture sectors. The refined sugar sector needs skilled manpower such as operations managers, production engineers, quality control managers, laboratory staff and logistics workforce. Jobs for steel fabricators, design engineers and steel erectors are created in the fabricated and structural materials sector. Plant, machinery and

Table 1 Relative changes 2(x1i − x0i )/(x1i + x0i ) and absolute changes (x1i − x0i ), with x0i being total output of sector i before balancing, and x1i the same after balancing. Sector

Gasoalcohol Alcohol Sugarcane Refined sugar Textile and canvas bags Agricultural services Business services Non-metallic mineral products Industrial machinery and equipment Fabricated metal products Raw sugar Paper products Industrial machinery repairs Data processing services Chemical products Basic chemicals Storage Water transport Wholesale trade Petroleum and coal products Crude oil Petrol and diesel

Relative change in total output (%)

Absolute change in total output (AU$m)

NA NA 74.95 1.48 0.75 0.29 0.22 0.16 0.16 0.15 0.12 0.09 0.08 0.08 0.07 −0.11 −0.13 −0.17 −0.21 −0.23 −1.93 −5.09

2557 639 259 10 1.6 0.6 7.1 1.9 12 11 3 1.5 0.3 1.2 3.7 −4.6 −12 −10 −1.1 −10 −320 −634

equipment industries need mechanical and electrical engineers to manage the infrastructure of the alcohol refinery. More drivers and logistics staff are needed in the road freight sector to transport feedstock to the alcohol refinery (Felix et al., 2007). The textile and canvas bags sector needs textile labourers to meet the increase in demand of canvas bags. The largest job losses are observed in the petrol and diesel sector. Due to a 20% reduction in the demand for petrol, fewer workers are needed for refining crude oil. The rail and water freight sector requires less train drivers, captains and officers to transport coal and crude oil, respectively. Interestingly, only about 130 jobs are lost in the crude oil sector, although the output decrease is substantial. This is due to the high wages ($160,000/year) paid in this sector, resulting in a low employment intensity (0.42FTE/$m). Coal, petroleum and gas sector experiences job losses of geoscientists, mining engineers and logistics personnel. Accountancy and finance jobs are lost in the financial services sector. The business management services sector loses administrative staff, consultants, typists and mail officers. Fewer technicians are needed in the machines and equipment repairing sector. Jobs are also lost in the hospitality sector due to a decrease in business trips and conferences. 4.3. Test of the Numerical Modified-RAS Balancing Approach It is interesting to study the changes in employment as the economy gradually adjusts to the introduction of the new sector. The sequential analysis and understanding of the stepwise adaptations occurring in the economy at various time intervals is reminiscent of Sequential Interindustry Model (SIM) by Romanoff and Levine (1981). Understanding the production chronologies resulting from a change in final demand is important if decision-makers need to take into account aspects of time-phased production into their planning of the scheduling of production processes (Romanoff and Levine, 1986). Before we apply our modified, column-only RAS approach, we first establish a) that the modified iterative algorithm converges, and b) that the final result is equal to that of the analytical approach. Fig. 4 shows the IO imbalance log10{|(xi)t − zi|, |(xp)t − zp|}t as a function of the adjustment stages of the RAS algorithm. The largest adjustments occur at early stages and the imbalance decreases after subsequent iterations. Individual sectors show different convergence behaviour, but

90

A. Malik et al. / Ecological Economics 107 (2014) 84–93

Fig. 3. Top-and bottom-ranking sectors in terms of changes in their employment (see Eq. 16) as a result of replacing petrol with alcohol. Positive values indicate gains in employment, whereas negative numbers signify job losses. Gasoalcohol is not included in the figure because, as explained in Section 2.4, no employment is allocated to this sector.

nevertheless the procedure converges, in our case, after about 30 iterations. We verified at this stage that the resulting input–output system ^ 1 determined using Eq. (9). was identical to A1 x The adjustment process alternates between industries and products (see the alternating pattern in Fig. 4). This is a consequence of the impossibility of simultaneous balancing industries and products in the augmented supply–use table, as described at the beginning of Section 2.1.1. In the first step inputs and output of each industry are balanced (which is reflected by the upper part of the first column being black) but demand and supply for the products are not (which is reflected by the lower part of the first column being light grey). In the second step it is exactly the other way around. Further steps continue this alternating behaviour but the light grey becomes darker (indicating less imbalances) and eventually becomes black (indicating negligible imbalances).

4.4. Employment Scheduling Requirements We use the numerical modified-RAS approach to enumerate the number of people required at subsequent adjustment stages of the economy. We understand the succession of employment requirements as resulting out of a planning process that anticipates only ever immediate suppliers. For example, after the introduction of alcohol and gasoalcohol production facilities, additional personnel is being recruited for sugarcane growing and sugar refining, but not yet for manufacturing additional agricultural equipment that may be needed to produce the additional cane. Similarly, after the reduction of petrol production, personnel is being laid off in the crude oil sector, but not yet for transporting the crude oil by ship. Thus the sequence of employment requirements across adjustment stages can be understood as a chronological succession. Immediately after the introduction of the gasoalcohol and alcohol sectors, many sectors experience an immediate change in employment (Table A1), and five of these sectors are large: alcohol, sugarcane, petrol and diesel, refined sugar and crude oil. Of these, the employment changes in the alcohol and petrol sectors are pre-defined in our case study. As a result of these two changes, sugarcane, crude oil and refined sugar sectors undergo immediate changes. Job gains in the direct suppliers of alcohol refinery (such as sugarcane and refined sugar) are much greater than job losses in the crude oil sector (Fig. 3). However, we have not considered delayed indirect employment effects yet. Interestingly, many other sectors undergo changes that are more complex. To study the indirect delayed effects, we have plotted the change in full-time (FTE) equivalent employment   n ^ xn −x0 ΔQ ¼ q

Fig. 4. System imbalance log10{|(xi)t − zi|, |(xp)t − zp|}t as a function of subsequent iterates of the modified RAS algorithm described in Section 2.3. x-axis shows iteration number, y-axis sector number. The starting point of the algorithm (x = 1) is product imbalance, as described in Section 2.3. Grey shades represent the log10 of transaction values expressed in '000 Australian Dollars. After 30 iterations, imbalances of individual industries and products are below 1000 Australian Dollars.

ð18Þ

for eight selected categories as a function of the RAS adjustment stages n (Fig. 5). Positive FTE values indicate job gains, whereas negative values indicate job losses. We observe the greatest job gain (2200 FTE) in the first adjustment stage of the economy. Immediately, labour is needed to produce inputs for constructing and operating the infrastructure of the alcohol refinery. At adjustment stage 2, job losses in the supply chain of petrol take effect, as evident, for example, from a steep decrease in employment for the rail and water freight sectors. The industries listed in the graph account

A. Malik et al. / Ecological Economics 107 (2014) 84–93

91

Fig. 5. Change in employment of sectors in the alcohol and petrol supply chains as a function of consecutive adjustment stages in the RAS sequence. Positive values represent employment gains and negative values signify employment losses. The adjustment of total employment change is also shown in the inset.

for a net job loss of −700 after 5 adjustment stages. After subsequent adjustment stages, job gains and losses counteract each other and the final net employment change amounts to about + 1400 jobs. Taking the entire transformation of the economy into account, this net result indicates that a future biofuel industry will likely be employmentpositive for Australia. 5. Discussion and Conclusions An existing input–output table can be augmented with a new sector to study the consequences of introducing new products and/or industries into an economy. We investigate the case of a future biorefining industry in Australia by introducing Brazil's alcohol refining technology into the Australian IO table. In particular we quantify changes in economic output and employment in the Australian economy, and we elucidate the employment scheduling requirements of individual economic sectors. Our results indicate an increase in the output of sectors providing direct input into the alcohol refinery, and vice versa for the petrol refinery. Initially, about 2000 staff will be required for establishing the new biofuel industry. Over time, however, job losses in the petrol supply chain will take effect and overall job gains will decrease. We conclude that a future biofuel industry will be employmentpositive for Australia. Employment effects of biofuel production have been documented in the literature for other nations as well, for example, Thailand (Silalertruksa et al., 2012), Brazil (Cunha and Scaramucci, 2006) and the European Union (Neuwahl et al., 2008). The results of this study have implications for environmental impact assessment (EIA) reports prepared for the government, because such reports often enumerate the consequences of new projects for employment, in addition to environmental impacts. In some cases, comments are made pertaining to the scheduling of employment for meeting the new project's timelines. An example is an EIA for a second Sydney airport (PPK Environment and Infrastructure, 1997a,1997b). Since this study did not employ IO techniques, a problem arose in terms of inconsistencies in quantifying the indirect employment effects of the airport (Lenzen et al., 2003), see also (Shepherd and Ortolano, 1996). Results of an input–output-assisted EIA include indirect upstream employment effects, and thus differ markedly from conventional EIAs. Whilst we have focused our investigation on employment effects, the topic of land-use change also requires special attention. To sustain a biofuel industry, the output of the sugarcane sector must increase by more than 70%. Currently, about 545,000 ha of land are used for producing $2 billion worth of cane per year (Cane Growers, 2010; DAFF, 2006).

To accommodate the new biofuel sector, another 71,000 ha of land will need to be allocated for growing feedstock. This additional land would potentially be made available in Queensland, since this region is the primary producer of sugarcane. Land-use change considerations were beyond the scope of our analysis. However, many studies have discussed this issue in detail (Marinoni et al., 2012). Searchinger et al. (2008) claim that diverting croplands for biofuel production releases carbon stored in the vegetation and soil into the atmosphere, thus increasing greenhouse gas emissions. They propose the use of carbon-poor lands and waste products (Fargione et al., 2008) for producing ethanol. We anticipate that using sugarcane bagasse, which is the fibrous residue left after juice extraction, will not lead to significant land-use changes in Australia. Nevertheless, this issue requires future analysis to investigate the effects of sugarcane-based biofuel production on local ecosystem and environment. A prerequisite for a successful biofuel industry is the availability of ethanol fuel throughout Australia, and also the introduction of flex-fuel vehicles capable of running on that fuel. To encourage Australian users to buy flex-fuel vehicles, the government should consider vehicle replacement schemes such as the “Cash for Clunkers” program and the “Scrapper Scheme” in USA and UK, respectively (Kagawa et al., 2011). Australia has a long way to go to match the biofuel production and consumption efficiency of its Brazilian counterpart.

Acknowledgements This work was financially supported by the Australian Research Council through its Discovery Projects DP0985522 and DP130101293, and by the National eResearch Collaboration Tools and Resources project (NeCTAR) through its Industrial Ecology Virtual Laboratory.8 NeCTAR is an Australian Government project conducted as part of the Super Science initiative and financed by the Education Investment Fund. The authors thank Sebastian Juraszek for expertly managing the advanced computation requirements. Special thanks go to Sangwon Suh from Bren School, University of California, Santa Barbara for his suggestions regarding hybrid LCA methods. We acknowledge the advice of Roberto Schaeffer from the Energy Planning Program at (COPPE/UFRJ) on the production recipe of alcohol. We also thank the reviewers for their insightful comments.

8

http://www.isa.org.usyd.edu.au/ielab/ielab.shtml.

92

A. Malik et al. / Ecological Economics 107 (2014) 84–93

Appendix A Table A1 Employment numbers (FTE) at various adjustment stages of the RAS algorithm aggregated over groups of similar sectors. Adjustment stages Sector Alcohol Sugarcane Refined sugar Plant, machinery and equipment Fabricated and structural materials Road freight Agriculture Textile and canvas bags Electricity supply Raw sugar Fertilisers and chemical products Wood, printing and publishing Fishing Forestry Furniture Insurance Miscellaneous manufacturing Miscellaneous metal products Refinery products Miscellaneous equipment Household appliances Personal services Services to mining Other textiles and clothing Food, beverages, tobacco Other heavy machinery Technical services Business management services Hospitality Financial services Crude oil Other remaining industries Coal, petroleum and gas Machines and equipment repairing Rail and water freight Petrol and diesel Total

0

1

2

3

4

5

6

7

∞

2223.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 −471.9 1751.7

2223.5 776.3 127.5 183.2 106.2 91.4 28.0 31.2 20.9 6.0 7.0 5.4 0.3 0.1 −0.4 0.9 −1.3 4.1 −3.1 −0.4 −6.4 −7.0 0.2 3.9 −8.8 −6.4 −20.7 −53.7 −53.8 −127.5 −132.4 −51.8 −127.7 −77.4 −268.6 −471.9 2196.7

2223.6 776.9 127.6 129.8 97.4 75.3 34.3 31.2 12.3 7.4 7.5 4.1 0.2 0.0 −0.5 −1.6 −1.5 0.2 −4.3 −3.2 −6.6 −7.3 −9.9 −1.6 −9.6 −14.2 −28.8 −54.2 −57.0 −128.3 −133.0 −96.0 −169.0 −237.1 −357.6 −472.1 1734.2

2223.5 777.0 127.6 112.6 90.1 68.3 35.8 31.1 10.0 7.6 7.5 3.1 0.2 0.0 −0.6 −1.7 −2.1 −1.3 −4.8 −4.8 −7.3 −7.6 −10.1 −8.1 −11.7 −16.9 −35.3 −54.5 −59.0 −128.9 −133.8 −119.8 −189.7 −286.7 −378.1 −472.3 1559.0

2223.4 777.0 127.5 106.8 86.6 64.2 35.8 31.1 9.0 7.7 7.4 2.5 0.2 −0.1 −0.7 −1.8 −2.4 −2.1 −5.1 −5.6 −7.5 −7.8 −10.2 −11.2 −13.1 −17.5 −38.5 −54.6 −59.8 −129.2 −134.0 −131.3 −198.8 −311.1 −386.7 −472.4 1477.4

2223.4 776.9 127.5 104.3 85.0 62.2 35.5 31.1 8.5 7.7 7.3 2.2 0.2 −0.1 −0.8 −1.8 −2.6 −2.6 −5.2 −6.0 −7.6 −7.9 −10.3 −12.6 −13.9 −17.8 −40.0 −54.7 −60.2 −129.3 −134.1 −137.0 −203.1 −322.7 −390.3 −472.5 1438.9

2223.4 776.9 127.5 103.0 84.3 61.1 35.3 31.1 8.2 7.7 7.3 2.1 0.1 −0.1 −0.8 −1.8 −2.7 −2.8 −5.2 −6.2 −7.7 −7.9 −10.3 −13.3 −14.3 −17.9 −40.7 −54.7 −60.4 −129.4 −134.2 −139.8 −205.2 −328.2 −391.9 −472.5 1420.2

2223.4 776.9 127.5 102.4 84.0 60.6 35.2 31.1 8.1 7.7 7.3 2.0 0.1 −0.1 −0.8 −1.9 −2.8 −2.9 −5.2 −6.3 −7.7 −7.9 −10.3 −13.6 −14.5 −18.0 −41.1 −54.7 −60.5 −129.4 −134.2 −141.1 −206.2 −330.9 −392.6 −472.5 1411.2

2223.4 776.9 127.5 101.9 83.6 60.2 35.1 31.1 8.0 7.7 7.2 2.0 0.1 −0.1 −0.8 −1.9 −2.8 −3.0 −5.3 −6.3 −7.7 −7.9 −10.3 −13.9 −14.7 −18.0 −41.4 −54.7 −60.6 −129.4 −134.2 −142.8 −207.1 −333.4 −393.4 −472.5 1402.5

References ABS, 2012. Australian National Accounts, Input–output Tables (Product Details), 2008–09. Australian Bureau of Statistics, Canberra, Australia. ABS, 2013. 9208.0 — Survey of Motor Vehicle Use, Australia, 12 months ended 30 June 2012, (Available from: http://www.abs.gov.au/ausstats/[email protected]/mf/9208.0, [accessed date: 11 June, 2013]). ACCC, 2013. Monitoring of the Australian Petroleum Industry — Report of the ACCC Into the Prices, Costs and Profits of Unleaded Petrol in Australia 2011: Chapter 6. Biofuels Australian Competition & Consumer Commission, Canberra, Australia. ACIL Tasman, 2009. Petroleum Import Infrastructure in Australia. Department of Resources, Energy and Tourism, Canberra, Australia. Acquaye, A.A., Wiedmann, T., Feng, K., Crawford, R.H., Barrett, J., Kuylenstierna, J., Duffy, A. P., Koh, S.L., McQueen-Mason, S., 2011. Identification of ‘carbon hot-spots’ and quantification of GHG intensities in the biodiesel supply chain using hybrid LCA and structural path analysis. Environ. Sci. Technol. 45, 2471–2478. AIP, 2013. AIP Annual TGP Data. Australian Institute of Petroleum, Canberra, Australia. Andreosso-O'Callaghan, B., Yue, G., 2000. An analysis of structural change in China using biproportional methods. Econ. Syst. Res. 12, 99–111. Bacharach, M., 1970. Biproportional Matrices and Input–Output Change. Cambridge University Press. Barros, Sergio, 2012. Global Agricultural Information Network (GAIN). Brazil Sugar Annual Report USDA Foreign Agricultural Service Washington, DC. Cane Growers, 2010. About the Australian Sugarcane Industry. Available from: http:// www.canegrowers.com.au/page/Industry_Centre/about-the-australian-sugarcaneindustry. Coordenação-Geral de Açúcar e Álcool, 2011. Mistura carburante automotiva (etanol anidro/gasolina) - cronología. Departamento da Cana-de-açúcar e Agroenergía, Secretaria de Produção e Agroenergía, Ministério da Agricultura, Pecuária e Abastecimento. Cunha, M.P., Scaramucci, J.A., 2006. Bioethanol as basis for regional development in Brazil: an input–output model with mixed technologies. 46th Congress of the European Regional Science Association (ERSA) (Volos, Greece).

DAFF, 2006. Sugar Industry Oversight Group Strategic Vision. Department of Agriculture, Fisheries and Forestry, Canberra, Australia. DAFF, 2012. Crops: Sugar, (Available from: http://www.daff.gov.au/agriculture-food/ crops/sugar, [accessed date: 12 June, 2013]). Dietzenbacher, E., Hoekstra, R., 2002. The RAS Structural Decomposition Approach. In: Hewings, G.J.D., Sonis, M., Boyce, D. (Eds.), Trade, Networks and Hierarchies: Modeling Regional and Interregional Economies. Springer-Verlag, Heidelberg, Germany. Dilekli, N., Duchin, F., 2014. Prospects for cellulosic biofuel production in the northeast: a scenario analysis. J. Ind. Ecol. (accepted for publication). Dobrescu, E., Gaftea, V., 2012. On the accuracy of RAS method in an emergent economy. Amfiteatru Econ. 14, 502–521. Duchin, F., 2004. Input–Output Economics and Material Flows. Rensselaer Polytechnic Institute, Troy, NY, USA. Duchin, F., Lange, G.-M., 1992. Technological choices and prices, and their implications for the US economy, 1963–2000. Econ. Syst. Res. 4 (1), 53–76. Duchin, F., Levine, S.H., 2011. Sectors may use multiple technologies simultaneously: the rectangular choice-of-technology model with binding factor constraints. Econ. Syst. Res. 23, 281–302. EEC, 2008. Eurostat Manual of Supply, Use and Input–output Tables. Office for Official Publications of the European Communities, Luxembourg. Fargione, J., Hill, J., Tilman, D., Polasky, S., Hawthorne, P., 2008. Land clearing and the biofuel carbon debt. Science 319, 1235–1238. Felix, E., Cardona, C.A., Quintero, J.A., 2007. Technical and Economic Viability of Biofuel Production Chains, Biofuels. Agriculture and Poverty Reduction Overseas Development Institute in Natural Resources Perspective. Gilchrist, D., St. Louis, L., 2004. An algorithm for the consistent inclusion of partial information in the revision of input–output tables. Econ. Syst. Res. 16, 149–156. Giljum, S., Hubacek, K., 2004. Alternative approaches of physical input–output analysis to estimate primary material inputs of production and consumption activities. Econ. Syst. Res. 16, 301–310. Green Car Congress, 2013. GM Begins Saab E85 Trial in Australia, Available from: http:// www.greencarcongress.com/2007/01/gm_begins_saab_.html, [accessed date: 14 June, 2013].

A. Malik et al. / Ecological Economics 107 (2014) 84–93 Guilhoto, J.J., Sesso Filho, U.A., 2005. Estimação da matriz insumo-produto a partir de dados preliminares das contas nacionais. Guilhoto, J.J.M., Sesso Filho, U.A., 2010. Estimação da matriz insumo-produto utilizando dados preliminares das contas nacionais: aplicação e análise de indicadores econômicos para o Brasil em 2005. Heijungs, R., Suh, S., 2002. The Computational Structure of Life Cycle Assessment. Springer. Hoekstra, R., van den Bergh, J.C., 2006. Constructing physical input–output tables for environmental modeling and accounting: framework and illustrations. Ecol. Econ. 59, 375–393. IBGE, 2011. Sistema de Contas Nacionais - Tabelas. Completas Instituto Brasileiro de Geografia e Estatística, Ministério de Planejamento e Orçamento, Rio de Janeiro, Brazil. Islam, M.R., Chhetri, A.B., Khan, M.M., 2012. Green Petroleum: How Oil and Gas Can Be Environmentally Sustainable. John Wiley & Sons, Canada. Joshi, S., 1999. Product environmental life-cycle assessment using input–output techniques. J. Ind. Ecol. 3, 95–120. Junius, T., Oosterhaven, J., 2003. The solution of updating or regionalizing a matrix with both positive and negative entries. Econ. Syst. Res. 15, 87–96. Kagawa, S., Nansai, K., Kondo, Y., Hubacek, K., Suh, S., Minx, J., Kudoh, Y., Tasaki, T., Nakamura, S., 2011. Role of motor vehicle lifetime extension in climate change policy. Environ. Sci. Technol. 45, 1184–1191. Kumar, S., Singh, N., Prasad, R., 2010. Anhydrous ethanol: a renewable source of energy. Renew. Sust. Energ. Rev. 14, 1830–1844. Lave, L.B., Hendrickson, E.C.C.T., McMichael, F.C., 1995. Using input–output analysis to estimate economy-wide discharges. Environ. Sci. Technol. 29, 420A–426A. Lenzen, M., Rueda-Cantuche, J.M., 2012. A note on the use of supply–use tables in impact analyses. Stat. Oper. Res. Trans. 36, 139–152. Lenzen, M., Murray, S.A., Korte, B., Dey, C.J., 2003. Environmental impact assessment including indirect effects—a case study using input–output analysis. Environ. Impact Assess. Rev. 23, 263–282. Lenzen, M., Gallego Luxan, B., Wood, R., 2009. Matrix balancing under conflicting information. Econ. Syst. Res. 21 (1), 23–44. Lenzen, M., Kanemoto, K., Moran, D., Geschke, A., 2012. Mapping the structure of the world economy. Environ. Sci. Technol. 46 (15), 8374–8381. Lenzen, M., Moran, D., Kanemoto, K., Geschke, A., 2013. Building Eora: a global multiregion input–output database at high country and sector resolution. Econ. Syst. Res. 25, 20–49. Lenzen, M., Geschke, A., Wiedmann, T., Lane, J., Anderson, N., Baynes, T., Boland, J., Daniels, P., Dey, C., Fry, J., Hadjikakou, M., Kenway, S., Malik, A., Moran, D., Murray, J., Nettleton, S., Poruschi, L., Reynolds, C., Rowley, H., Ugon, J., Webb, D., West, J., 2014. Compiling and using input–output frameworks through collaborative virtual laboratories. Sci. Total Environ. 485, 241–251. Leontief, W., Ford, D., 1970. Environmental repercussions and the economic structure: an input–output approach. Rev. Econ. Stat. 262–271. Li, X., Feng, K., Siu, Y.L., Hubacek, K., 2012. Energy-water nexus of wind power in China: the balancing act between CO2 emissions and water consumption. Energy Policy 45, 440–448. Liu, C.H., Lenzen, M., Murray, J., 2012. A disaggregated emissions inventory for Taiwan with uses in hybrid input–output life cycle analysis (IO-LCA). Nat. Res. Forum 36, 123–141. Macedo, I.C., Seabra, J.E., Silva, J.E., 2008. Green house gases emissions in the production and use of ethanol from sugarcane in Brazil: the 2005/2006 averages and a prediction for 2020. Biomass and Bioenergy 32, 582–595. Marinoni, O., Navarro Garcia, J., Marvanek, S., Prestwidge, D., Clifford, D., Laredo, L., 2012. Development of a system to produce maps of agricultural profit on a continental scale: an example for Australia. Agric. Syst. 105, 33–45. Miller, R.E., Blair, P.D., 2009. Input–Output Analysis: Foundations and Extensions, 2 ed. Cambridge University Press, Cambridge.

93

Neuwahl, F., Löschel, A., Mongelli, I., Delgado, L., 2008. Employment impacts of EU biofuels policy: combining bottom-up technology information and sectoral market simulations in an input–output framework. Ecol. Econ. 68, 447–460. NZIC, 2008. Refining Crude Oil, (Available from: http://nzic.org.nz/ChemProcesses/ energy/7A.pdf, [accessed date: 3 July, 2013]). Olbrich, H., 1963. The Molasses. Institut für Zuckerindustrie, Berlin, Germany. PPK Environment and Infrastructure, 1997a. Draft Environmental Impact Statement — Second Sydney Airport Proposal. Commonwealth Department of Transport and Regional Development, Canberra, Australia. PPK Environment and Infrastructure, 1997b. Summary of the Environmental Impact Statement for the proposed Second Sydney Airport at Badgerys Creek. Commonwealth Department of Transport and Regional Development, Canberra, Australia. Romanoff, E., Levine, S.H., 1981. Anticipatory and responsive sequential interindustry models. IEEE Trans. Syst. Man Cybern. 11, 181–186. Romanoff, E., Levine, S.H., 1986. Capacity limitations, inventory, and time-phased production in the sequential interindustry model. Pap. Reg. Sci. 59, 73–91. Rose, A., 1984. Technological change and input–output analysis: an appraisal. Socio Econ. Plan. Sci. 18, 305–318. Schmitz, A., Wilson, N.L., Mass, C.B., Zilberman, D., 2011. The Economics of Alternative Energy Sources and Globalization. Bentham Science Publishers. Searchinger, T., Heimlich, R., Houghton, R.A., Dong, F., Elobeid, A., Fabiosa, J., Tokgoz, S., Hayes, D., Yu, T.-H., 2008. Use of US croplands for biofuels increases greenhouse gases through emissions from land-use change. Science 319, 1238–1240. Shepherd, A., Ortolano, L., 1996. Strategic environmental assessment for sustainable urban development. Environ. Impact Assess. Rev. 16, 321–335. Silalertruksa, T., Gheewala, S.H., Hünecke, K., Fritsche, U.R., 2012. Biofuels and employment effects: implications for socio-economic development in Thailand. Biomass and Bioenergy 46, 409–418. Stone, R., 1961. Input–Output and National Accounts. Organisation for Economic CoOperation and Development (OECD), Paris, France. SugarCane.org, 2013. Virtual Tour of a Sugarcane Mill (Transcript). Available from: http:// sugarcane.org/about-sugarcane/virtual-mill-tour/virtual-tour-of-a-sugarcane-milltranscript ([accessed date: 2 July, 2013]). Suh, S., 2004. Functions, commodities and environmental impacts in an ecological– economic model. Ecol. Econ. 48, 451–467. Sun, Y., Cheng, J., 2002. Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresour. Technol. 83, 1–11. Tewari, J.C., Malik, K., 2007. In situ laboratory analysis of sucrose in sugarcane bagasse using attenuated total reflectance spectroscopy and chemometrics. Int. J. Food Sci. Technol. 42, 200–207. Tukker, A., de Koning, A., Wood, R., Hawkins, T., Lutter, S., Acosta, J., Rueda Cantuche, J.M., Bouwmeester, M., Oosterhaven, J., Drosdowski, T., 2013. EXIOPOL—development and illustrative analyses of a detailed global MR EE SUT/IOT. Econ. Syst. Res. 25, 50–70. van der Linden, J.A., Dietzenbacher, E., 1995. The Nature of Changes in the EU Cost Structure of Production 1965–1985: An RAS Approach. In: Armstrong, H., Vickerman, R. (Eds.), Convergence and Divergence among European Regions, London, Pion, pp. 124–139. van der Linden, J.A., Dietzenbacher, E., 2000. The determinants of structural change in the European Union: a new application of RAS. Environ. Plan. A 32, 2205–2230. Wiedmann, T.O., Suh, S., Feng, K., Lenzen, M., Acquaye, A., Scott, K., Barrett, J.R., 2011. Application of hybrid life cycle approaches to emerging energy technologies—the case of wind power in the UK. Environ. Sci. Technol. 45, 5900–5907. X-RATES, 2009. Monthly Average, (Available from: http://www.x-rates.com/average/? from=BRL&to=AUD&amount=1.00&year=2009, [accessed date: 7 August, 2013].). Zhu, J., Gleisner, R., Scott, C., Luo, X., Tian, S., 2011. High titer ethanol production from simultaneous enzymatic saccharification and fermentation of aspen at high solids: a comparison between SPORL and dilute acid pretreatments. Bioresour. Technol. 102, 8921–8929.