Assessing the eco-efficiency of U.S. manufacturing industries with a focus on renewable vs. non-renewable energy use: An integrated time series MRIO and DEA approach

Journal Pre-proof Assessing the eco-efficiency of U.S. manufacturing industries with a focus on renewable vs. non-renewable energy use: An integrated time series MRIO and DEA approach Bahadir Ezici, Gokhan Egilmez, Ridvan Gedik PII:

S0959-6526(19)34500-7

DOI:

https://doi.org/10.1016/j.jclepro.2019.119630

Reference:

JCLP 119630

To appear in:

Journal of Cleaner Production

Received Date: 1 June 2019 Revised Date:

14 November 2019

Accepted Date: 8 December 2019

Please cite this article as: Ezici B, Egilmez G, Gedik R, Assessing the eco-efficiency of U.S. manufacturing industries with a focus on renewable vs. non-renewable energy use: An integrated time series MRIO and DEA approach, Journal of Cleaner Production (2020), doi: https://doi.org/10.1016/ j.jclepro.2019.119630. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

Assessing the Eco-Efficiency of U.S. Manufacturing Industries with a Focus on Renewable vs. Non-Renewable Energy Use: An Integrated Time Series MRIO and DEA Approach 1

Bahadir Ezici, 2Gokhan Egilmez, 3Ridvan Gedik

1

Dept. of Mechanical and Industrial Engineering, University of New Haven, West Haven, CT

2

Department of Mechanical and Industrial Engineering, University of New Haven, West Haven, CT

(Corresponding Author, Email: [email protected]) 3

Department of Management, Quinnipiac University, Hamden, CT

Abstract This study investigated the global supply chain-linked renewable and nonrenewable energy use impacts and economic output of U.S. manufacturing industries over a twenty-year period. Considering energy use impacts and economic outputs together within the scope of a global-trade-linked, cradle-to-gate life cycle provides a comprehensive understanding of the environmental and economic impacts of industrial activities. In the first phase of the methodology, twenty multi-region input-output models were built to assess the energy and economic output nexus using a time series approach. Sixteen energy carriers were considered and aggregated in terms of renewable and nonrenewable energy use impacts. The second phase of the methodology focused on benchmarking U.S. manufacturing industries’ eco-efficiency, considering renewable to nonrenewable energy use and economic output. To accomplish this, data envelopment analysis (DEA) models were developed, and two benchmarking (eco-efficiency) measures were proposed, namely: renewability ratio (RR) and economic-output-induced renewability ratio (E-RR). The results indicated that the economic output of the manufacturing industries exhibited a steady, sustainable growth. Similar growth was observed in nonrenewable energy use. In contrast, the trend in renewable energy use was seen to be stagnant. No statistically significant improvement was observed in either the RR or E-RR measures, which were found to parallel the multi-region input-output analysis

1

(MRIO) results. Although an increase in both the mean RR (0.3 to 0.4) and the mean E-RR (0.38 to 0.52) scores was recorded from 1995 to 2014, this was concluded to be unsatisfactory, since the majority of the industries’ eco-efficiency results were still below 0.5. Such an unsatisfactory result could be attributed to an imbalanced growth in nonrenewable energy use and economic output relative to renewable energy use. The findings of this study suggest that substantial policy changes are required immediately to shift the negative trend in renewable energy use to comply with UN Sustainability Development Goals 7 and 13. Key Words: renewable energy; renewability ratio; manufacturing; sustainable development; multi-region input-output (MRIO) analysis; data envelopment analysis (DEA)

2

1. Introduction Manufacturing industries play a prime role in the sustainable growth of a country’s economy. The economic output of U.S. manufacturing industries has been growing steadily since the 2008 financial crisis (BEA – Gross Output of United States (U.S.) Manufacturing, 2018). In addition, U.S. manufacturing industries employ nearly 12.5 million workers, accounting for 8.5 % of the overall workforce. With the continuous economic growth of manufacturing industries come energy and natural resource supply challenges, as well as environmental stresses, such as increased carbon, energy, material, and land footprints. In terms of greenhouse gas (GHG) emissions, U.S. manufacturing industries contribute the third largest portion, a 22% share, after electricity generation (28%), and transportation (29%) (EPA, 2019). In this context, since energy consumption and GHG emissions are highly correlated – as U.S. energy sources tend to be more nonrenewable (Kucukvar et al., 2016; Egilmez et al., 2013) – assessing the relationships between manufacturing industries and their dependency on renewable and nonrenewable energy is a crucially important undertaking for the effective transformation of the U.S. economy into a sustainable one in all areas of sustainability. The EIA's recently released energy outlook (International Energy Outlook, 2013) projects that energy consumption in the world will grow by 56% between 2010 and 2040. According to the data, global energy consumption is still highly dependent on fossil fuels. In 2012, fossil fuels accounted for 84% of worldwide energy consumption (U.S. Energy Information Administration, 2016), and these statistics are similar to those for the U.S. In fact, although the U.S. has been a highly industrialized country for a long time (e.g., the industrial sector accounted for about one-third of U.S. energy consumption in 2016), fossil fuels are still the largest source of energy for electricity generation. In 2016, About 63% of electricity generation was from nonrenewable energy sources, namely, fossil fuels (i.e., coal, natural gas, petroleum, and other gases); about 20% was generated from nuclear energy, and about 17% from renewable energy sources (U.S. Energy Information Administration, 2018).

3

Studying the renewable vs. nonrenewable energy use of its manufacturing industries is important for establishing effective sustainable development plans and policy making in the U.S. Thus, it is evident that the energy sourcing of manufacturing activities in the U.S. is crucially important from economic, social, and environmentally sustainable (e.g., carbon, energy, and water footprint) perspectives. Even though recent outlook reports identify the manufacturing industries’ share of energy consumption in the U.S. economy, it is also important to study this problem from a global supply chain perspective, due to the critical share (27%) of the U.S. Gross Domestic Product (GDP) in international trade (The World Bank, 2018). Therefore, this study primarily aimed to investigate the economic output and renewable versus nonrenewable energy use impacts of U.S. manufacturing industries, considering international trade linkages between 1995 and 2014, and its secondary objective was to assess the industrial eco-efficiency performance of these industries, with a focus on renewable energy use over the study period.

2. Literature Review The proposed approaches in this study consist of an input-output analysis (IOA) and renewable energy-focused eco-efficiency assessment methods. Thus, the relevant literature is reviewed in two sections corresponding to these. In the last part of the review, the research gap and the contributions of the current study are described.

2.1. Applications of Input-Output Analysis (IOA) to Energy Use Projection Among the methods used to conduct life-cycle assessment (LCA), input-output analysis (IOA) has become a robust and widely-applied approach, since it does not require extensive and exhaustive data collection, and it is effective for studying industry-wide sustainability impacts (Suh et al., 2004). Typically, it enables expanded system boundaries (it can include regional, national, or international supply chains) and provides a robust framework for studying any type of process and/or product whose production or service activities can be expressed in terms of monetary values and entered into an IO-LCA model (Miller & Blair, 2013). Even though IO-LCA has been understood as environmentally extended

4

form of input-output analysis (IOA), today we find it being applied to the social, economic, environmental, and ecological dimensions of sustainable development. Over the last two decades, IO-LCA applications have been widely used to study the environmental, economic, and social impacts of industrial processes on regional or national economic scales (Park et al., 2016). The relevant literature abounds with works that propose IO-LCA approaches for investigating the direct and indirect (supply chain-linked) environmental, social, and economic implications of products, processes, and industries. For instance, Wiebe et al. (2012) studied the direct and indirect impact energy use and CO2 emissions associated with the production of goods and services in 53 countries, 2 regions, and 48 sectors. Pan et al. (2017) explored the relationships between energy supply and demand in order to provide an optimal design for energy management. Palmer (2017) developed an environmentally extended IOA to estimate energy flows. Wu and Chen (2017) compared China`s energy use with the rest of the world, using an extended IOA approach. Hamilton and Kelly (2017) studied the interactions among the energy, water, and food impacts of products at different levels of supply chains. Chen et al. (2018) investigated the regional, national, and global levels of embodied energy flow networks by adapting the environmentally extended IOA. Even though IO-LCA approaches offer credible advantages in terms of tracing direct and indirect (supply chain) impacts at regional or national economic levels, previous works mainly employed a singleregion IO-LCA approach, based on a domestic technology assumption. This creates critical limitations to tracing sustainability impacts at the global economic level (Park et al., 2016). Therefore, the MRIO approach has recently been adopted to overcome the limitations of single-region IO-LCA approaches (Andrew et al. (2009)). Wiedmann et al. (2011) emphasized why the MRIO analysis has become the mainstream IOA approach nowadays, listing and discussing the advantages of the framework and suggesting future work on the model. Some of the advantages of MRIO include the following: (i) it provides the ability to trace impacts associated with onsite, domestic and global supply chain-level activities, which overall typically involves multiple industries and multiple countries, (ii) it can be

5

extended for forecasting and projection applications and, (iii) it provides a platform for accurately studying the impacts of complex products and processes that have global supply chains. In terms of the applications of MRIO, Wiedmann (2009) compared the single- and multi-region IO models and analyzed work on tracing consumption-based emissions and resource accounting. Moreover, Lenzen et al. (2010) proposed the use of a MRIO analysis to investigate uncertainty (standard deviation) in regard to carbon multipliers in the United Kingdom’s economy. Su and Ang (2014) traced the effects of inter-regional and international trade on the carbon footprint of China. Zhang et al. (2015) divided China into seven regions and studied energy flows within and across these regions, using a MRIO modeling approach. Zhang et al. (2016) investigated energy transfers in China, based on geographical and time changes, using MRIO models for domestic trading in specific time series. Hong et al. (2016) researched the energy use associated with consumption and inter-regional trade in the construction industry within China. Xia et al. (2017) investigated coal routes and coal utilization in the world, using a MRIO analysis for a set of coal species. Sun et al. (2017) studied three major regions in China and their energy consumption and contributions to sustainable development. Nakanoet al. (2018) proposed the use of a MRIO analysis to investigate a next-generation energy system (IONGES) related to renewable energy. Ali et al. (2018) analyzed carbon and water footprints in Italy within a MRIO framework. Zhang et al. (2013) used MRIO modeling to examine hidden energy flows domestically and inter-regionally in China. These works clearly indicate that MRIO analysis has been predominantly used for studying energy use and energy footprints in emerging and developed economies, and considering supply chain impacts within and between various regions of the world economy. In this regard, Table 1 provides a summary of relevant works that investigate carbon and energy footprints through the use of MRIO analysis. Carbon and energy use impacts are highly correlated with each other in many regions of the world, due to a great dependency on fossil fuels for power production. In this regard, studying the energy use impacts of growing world economies is a matter of importance.

6

Table 1. Summary of the IOA-based Energy Footprint Assessment Literature Study

Focus

Scope

Method

1

Wiedmann (2009)

Energy, GHG emissions

Survey

IO & MRIO

2

Andrew et al. (2009)

Carbon footprint

87 Countries

MRIO

3

Lenzen et al. (2010)

CO2

UK

MRIO

4

Kanemoto et al. (2011)

Economy

187 Countries

MRIO

5

Wiedmann et al. (2011)

IOA Theory

Methods

MRIO

6

Wiebe et al. (2012)

Energy-related CO2

53 Countries, 2 Regions, 48 Sectors

MRIO

7

Zhang et al. (2013)

Energy

China

IO-MRIO

8

Su & Ang (2014)

Carbon

China

MRIO

9

Lindner & Guan (2014)

Carbon

China

EIOA

10

Y. Zhang et al. (2015)

Energy

China

MRIO

11

Rocco & Colombo (2016)

Energy

Global

IO-MRIO

12

Honget al. (2016)

Energy

China

MRIO

13

Hong et al. (2016)

CO2

Australia

MRIO

14

Kucukvar et al. (2016

Carbon, Energy

Global

MRIO

15

Chen et al. (2017)

Coal

Global Economy

MRIO

16

Wu & Chen (2017)

Energy

China

IO

17

Palmer (2017)

Energy

Australia

EE-IO

18

Pan et al. (2017)

Energy

China

IO

19

Hamilton & Kelly (2017)

CO2

Sub-Saharan Africa

IO

20

Sun et al. (2017)

Energy

China

MRIO

21

Owen et al. (2018)

Energy-Water

UK

IO & MRIO

22

Nakano et al. (2018)

Energy

Japan

MRIO

23

Chen et al. (2018)

Energy

Global

EEIOA

24

Ali et al. (2018)

CO2

Italy

MRIO

25

Kucukvar et al. (2019

GHG, Energy

Global

MRIO

The U.S. has been among the most sustainable and powerful economies, influencing global economic policy making and trade. In this regard, as in any country, manufacturing industries play a critical role in the economic output growth of the U.S. While providing enormous benefits to the economy and society through employment, manufacturing industries are also responsible for a considerable amount

7

of environmental impacts, specifically in the areas of greenhouse gas (GHG) emissions and energy use (Egilmez et al., 2017). Recent reports also indicate that energy used by industrial activities is among the top contributors to the overall energy and carbon footprint in the U.S. (U.S. Energy Information Administration, 2017). In this context, the majority of the relevant literature has addressed the economic and environmental impacts of U.S. manufacturing by using single-region IO-LCA models. For instance, five environmental impact categories were studied by Egilmez et al. (2013); the U.S. manufacturing and transportation nexus was studied by Egilmez and Park (2014); TRACI impacts were assessed by Park et al. (2015); an environmental impact assessment of U.S. agricultural production was conducted by Park et al. (2016); and a carbon footprint stock analysis of U.S. manufacturing was carried out by Egilmez et al. (2017). In a recent study, Kucukvar et al. (2016) developed MRIO models to study the global supply chain impacts of Turkish manufacturing industries. Similarly, Kucukvar et al. (2019) analyzed the sustainability impacts of the world’s largest food producers, using an MRIO framework. It was found that agriculture was the most dominant industry in food supply chains, and India, China, and Russia were found to have the highest energy use and carbon footprints per value added. Abbod (2016) investigated the multi-region carbon and energy use impacts of U.S. manufacturing industries from a stochastic MRIO perspective, in which energy use was an aggregation of all renewable and nonrenewable energy sources. While this work addressed the energy use impacts of U.S. manufacturing economic output on global supply chains, energy use was traced as a whole, and no specific attention was paid to the specific renewable and nonrenewable energy carriers; moreover, the study period was limited and out of date. Hence, it is crucial to study renewable and nonrenewable energy use impacts in detail, and this is the primary focus of this study.

2.2. Eco-Efficiency Assessment with Data Envelopment Analysis (DEA) Manufacturing industries use both nonrenewable and renewable energy sources to carry out their operations and produce products for other industries and customers. In this regard, the concept of eco-

8

efficiency concept has been widely applied to assess the economic contributions of specific industries, products, or services against their associated environmental impacts. Eco-efficiency is assessed in terms of the ratio of the economic benefits of a decision-making units (DMU’s) activity to the environmental impacts of the corresponding activity (Egilmez et al., 2013). Among the quantitative approaches, DEA is widely used for eco-efficiency analysis (Huppes & Ishikawa, 2005), due to its being a robust, linear, programming-based, benchmarking approach. DEA is typically used to compare the performance of DMUs (Sembill & Dreyer, 2009; Egilmez & Stewart, 2019). Moreover, DEA has been used to compare the performance of banks, hospitals, stock market companies, schools, universities, etc. (Egilmez et al., 2016). DMUs can take the form of countries, industries, higher institutions, and so on, where a DMU is to be benchmarked with the rest of the DMUs in a study sample. DEA experiments result in an efficiency score, typically between 0 and 1 for each DMU, which indicates relative performance based on the ratio of output (produced) to input (used) for a given production activity (Sherman & Zhu, 1997). In the literature, efficiency scores are typically based on the ratio of output(s) to input(s); thus, DMUs with greater output produced and less input used are considered efficient and yield a score of 1 (Park et al., 2016). Chien and Hu (2007) analyzed the impact of renewable energy use on the technical efficiency of 45 countries for a specific time period. Park et al. (2016) integrated DEA with ecologically based lifecycle assessment (Eco-LCA) and analyzed the impacts related to agricultural and food production activities in the U.S. Furthermore, the literature abounds with studies that used DEA to examine the ecoefficiency of particular agricultural products. For instance, Vázquez-Rowe et al. (2012), combined LCA + DEA models conduct a performance analysis of grape producers in Spain, and concluded that the producers could increase their efficiency by reducing the material inputs (Chenel et al., 2014). Similarly, the farming of nectarines (Nabavi-Pelesaraei et al., 2014), tangerines (Qasemi-Kordkheili et al., 2014), and watermelons (Ashkan Nabavi-Pelesaraei et al., 2016) in Iran was studied with a DEA approach to examine its eco-efficiency in regard to energy consumption and GHG emissions. Other studies of agricultural industries looked at rice producers (Ashkan Nabavi-Pelesaraei et al., 2014) and orange producers (Ashkan Nabavi-Pelesaraei et al., 2014).

9

A number of studies have focused on the energy use efficiency of countries and industries using different methodologies. For example, Miketa and Mulder (2005) studied the energy-productivity performance of 56 countries in 10 manufacturing sectors during the period 1971-1995. Phylipsen et al. (1997) conducted an efficiency analysis of manufacturing energy intensity among countries, considering the structural differences among the countries. Ramanathan (2005) studied the energy consumption and carbon dioxide emissions of 17 countries in the Middle East and North Africa by adapting DEA and focusing on energy consumption and CO2 levels. Sözen and Alp (2009) used DEA to compare Turkey`s energy use efficiency with that of the European Union (EU) countries in terms of energy consumption, local pollutants, and GHG emissions. Honma and Hu (2014) estimated total-factor energy efficiency (TFEE) scores for 47 regions between 1996 and 2008 by adopting a stochastic frontier analysis model. Mukherjee (2008) applied DEA to analyze the energy efficiency of the aggregate manufacturing sectors. In recent work, Kouchaki-Penchah et al. (2017) applied a methodology that integrated DEA and LCA, whose objective was to determine the energy efficiency of tea production in Guilan province, Iran, and to help to reduce its environmental burdens. Nitrogen, diesel fuel, and machinery were identified as the main hotspots in the majority of impact categories, and the results showed that a 17% to 20% reduction in emissions would be required if inefficient farms were to be full eco-efficient. In another study closely related to this one, Hosseinzadeh-Bandbafha et al. (2017) employed DEA to quantify the energy efficiency of fattening farms. GHG emission reductions were distributed among the excessive energy consumption for the inefficient farms, which was the case for approximately 44% of the total number of farms studied. The use of stochastic frontier analysis (SFA) has also been reported in the literature in place of DEA. For instance, Shabanzadeh-Khoshrody et al. (2016) proposed a framework integrating the Tornqvists-Theil (TTP) index, SFA, and matching methods to investigate the effects of dam construction on the productivity and efficiency of farmers. Egilmez et al. (2013) used EIO-LCA to quantify the single-region economic output and five environmental impacts (GHG emissions, energy use, water use, toxic releases, and hazardous waste generation) of the U.S. manufacturing industries for a single-year study period, in 2007. In the latter part

10

of this work, eco-efficiency analysis was conducted by using economic output as the nominator and environmental impacts as the denominator. Even though this work is the study closest to the secondary focus of this paper, there were four major limitations: (1) The energy impacts were considered as a whole (renewable and nonrenewable), which did not enable the researchers to discern the separate impacts of renewable and nonrenewable energy carriers. (2) A single-region IO-LCA approach was used, which holds to the domestic technology assumption (Miller & Blaier, 2013), and thus, the global supply chain impacts of U.S. manufacturing activities were neglected. (3) The study used only data from the year 2007, which are quite outdated today, and this did not allow the researchers to study patterns of change over a longer time period. (4) The eco-efficiency analysis considered five environmental impacts together, which provided no insights into renewable versus nonrenewable energy use. Based on the IOA literature review, which was summarized in Table 1, there have been no studies on the renewable and nonrenewable energy impacts of U.S. manufacturing industries on a global scale. Therefore, this study investigated the economic output and the renewable and nonrenewable energy use impacts of U.S. manufacturing activities over a twenty-year study period. The scope of energy use involved energy use onsite (in production processes) and the national (domestic) and global supply chain industries. The secondary focus was to investigate the eco-efficiency of the industries, with a specific focus on renewable versus nonrenewable energy use.

3. Methodology In this section, the proposed MRIO and DEA approaches are explained in detail, as well as the data collection and preparation. The methodology consisted of the steps explained in the following sections and depicted in Figure 1. The results of Phase 2 (MRIO experimentation) were used as the input for Phase 3.

11

Phase 1 1. Data Collection 2. Data Preparation

Phase 2 1. MRIO Modelling 2. Experimentation 3. Analysis of Results

Phase 3 1. Data Normalization 2. DEA Modeling 3. DEA Experimentation 4. Analysis of Results

Figure 1. Hierarchical framework of the proposed methods. 3.1. Data Collection and Preparation The data were obtained from the World Input Output Database (2018), which covers the period from 1995 to 2014. This twenty-year study period was chosen based on the availability of data on economic input and output, final demand, and environmental impact multipliers. The term domestic transaction refers to the economic flow between pairs of industries in the same country. On the other hand, global transaction indicates the economic flows between pairs of industries located in different countries. The WIOD database contains information on the economies of 40 countries (see Table 3) and the Rest of the World (ROW). Each country’s economy is represented in terms of 35 industries (16 of them manufacturing) (see Table 4). The codes used by the WIOD for these industries and countries are shown in Tables 2 and 3, respectively (Dietzenbacher et al., 2013). Table 2. Primary energy carriers in the WIOD (Dietzenbacher et al., 2013) Primary Energy Carriers Crude oil Coal Natural Gas Nuclear Energy Renewable

WIOD Code Crude HCoal, BCoal, Coke NatGas, OthGas Nuclear Waste, Biogasol, Biodiesel, Biogas, Geotherm, Solar, Wind, Othsourc, Hydro

12

Table 3. WIOD countries and their abbreviations (Dietzenbacher et al., 2013) Codes USA CHN CAN SWE MEX IND BRA JPN KOR DEU TWN GBR AUS IDN FRA ITA NLD ESP POL BEL TUR FIN

Country United States China Canada Sweden Mexico India Brazil Japan Korea Germany Taiwan United Kingdom Australia Indonesia France Italy Netherlands Spain Poland Belgium Turkey Finland

Codes ROU CZE DNK RUS IRL AUT BGR HUN PRT GRC SVK EST LTU SVN LUX LVA CYP MLT RoW

Country Romania Czech Republic Denmark Russian Federation Ireland Austria Bulgaria Hungary Portugal Greece Slovakia Estonia Lithuania Slovenia Luxembourg Latvia Cyprus Malta Rest of The World

The energy consumption multipliers (TJs per $M economic output) for the analysis come from the WIOD as well. The energy use multiplier data consists of 16 main energy sources (carriers), which are categorized as renewable or nonrenewable energy sources, as shown in Table 2. Nuclear energy is grouped under nonrenewable energy carriers because the uranium consumed in its production is identified as nonrenewable by the U.S. Energy Information Administration, (2018). The units of measure for the energy sources vary by type: liquid fuels are represented in barrels or gallons, natural gas in cubic feet, coal in short tons, and electricity in kilowatts or kilowatt-hours (U.S. Energy Information Administration, 2018). In this study, energy data are represented using tera-joules (TJ), and monetary flow is represented in millions of dollars of economic activity (M$) (Kucukvar et al., 2016, 2019).

13

Table 4. WIOD manufacturing sectors and their abbreviations Industry Agriculture, Hunting, Forestry and Fishing Mining and Quarrying Food, Beverages and Tobacco Textiles and Textile Products Leather, Leather and Footwear Wood and Products of Wood and Cork Pulp, Paper, Paper, Printing and Publishing Coke, Refined Petroleum and Nuclear Fuel Chemicals and Chemical Products Rubber and Plastics Other Non-Metallic Mineral Basic Metals and Fabricated Metal Machinery, Nec. Electrical and Optical Equipment Transport Equipment Manufacturing, Nec; Recycling Electricity, Gas and Water Supply Construction Sale, Maint. and Repair of Motor Vehicles and Motorcycles; Retail Sale of Fuel Wholesale Trade and Commission Trade, Except of Motor Vehicles and Motorcycles Retail Trade, Exc. of Motor Vehicles and Motorcycles; Repair of Household Goods Hotels and Restaurants Inland Transport Water Transport Air Transport Other Supporting and Auxiliary Transport Activities; Activities of Travel Agencies Post and Telecommunications Financial Intermediation Real Estate Activities Renting of M&Eq and Other Business Activities Public Admin and Defence; Compulsory Social Security Education Health and Social Work Other Community, Social and Personal Services Private Households with Employed Persons

Abbreviations AHFF MQ FBT TTP LLF WPWC PPPPP CRPNF CCP RP ONMM BMFM MN EOE TE MNR EGW C SMRMVM WTCT RTEMVM HR IT WT AT OSATA PT FI REA RMOBA PAD E HSW OCSPS PHEP

Furthermore, the MRIO results were normalized prior to the modeling and experimentation of the proposed DEA approach, in order to eliminate the negative impacts of imbalances in the data due to the

14

use of different units such as TJs and millions of dollars. The mean normalization method was used because it has been used as the preferred normalization method in previous eco-efficiency assessment studies such as Vázquez-Rowe et al., 2012b; Iribarren & Vázquez-Rowe, 2013; and Egilmez et al., 2013. 3.2. The Proposed MRIO Framework

The MRIO approach was employed in order to quantify the onsite (production), national (domestic), and global supply chain-level impacts of the U.S. manufacturing industries’ economic output and their associated nonrenewable and renewable energy use. The MRIO models were developed using MATLAB software. They were run to estimate the energy use and economic output impacts of the manufacturing industries on a global scale (onsite + domestic supply chains + global supply chains). Then the energy use impacts were aggregated into renewable and nonrenewable energy use totals. Thus, intense energy-using industries were identified over the study period. Furthermore, a DEA approach was adopted to conduct the eco-efficiency analysis, which was aimed at identifying the efficiency of industries, based on their relative dependency on renewable energy. In this regard, two eco-efficiency measures are proposed. A simple standard IO model is given below in equation 1 (Miller & Blair, 2009):
= 
+ 

,

(1)

where A is the direct requirement matrix (transaction matrix), f is the final demand, and x is the total output. X can be termed
= , using the Leontief inverse ( = ( −   (Leontief, 1970). Equations 2, 3, and 4 indicate the inter-industry transactions matrix, the final demand vector, and the total industry output vector, respectively (Kucukvar et al., 2016).    =    

     

      

(2)

In equation 2, Z represents the monetary transaction matrix among all pairs of industries. The Z matrix is transformed into an A matrix to standardize the monetary flows between 0 and 1, as shown in equation 3 (Miller & Blair, 2013):

15

  =  

   

where  =  


(3)

.

The Leontief inverse (L) and the total requirement matrix are represented in equations 4 and 5 (Miller & Blair, 2009): −   −  =  − −

 =  − 



−  =  − −

− −   − − −   −

(4)



(5)

.

Thus, the total energy use (y) for a specific energy carrier can be derived as shown in equation 6, where y is column vector of total energy use, E is a diagonal matrix of direct energy use per milliondollars-worth of economic output (the impact multiplier), and f is the final demand.  = 

.

(6)

MRIO analysis is widely used to track the monetary flows of industries and countries. Furthermore, environmental impacts can be estimated using MRIO analysis at the domestic onsite, domestic supply chain, and international supply chain levels. In the set of equations above, the researchers implemented MRIO analysis for each of the 16 renewable and nonrenewable energy carriers, over the study period from 1995 to 2014, to estimate the total energy use (y) and the economic output (x). Once the renewable and nonrenewable energy use of the U.S. manufacturing industries was traced on a global scale with the above described framework, the results of the MRIO experiments were used as the input data for the DEA models and experiments. In this context, DEA was employed to evaluate the ecoefficiency performance of the manufacturing industries based on the type of energy they used (renewable versus nonrenewable) and the total economic output.

16

3.3. Proposed DEA Approach In this study, input-oriented DEA is used to conduct an energy-use efficiency assessment of industries and countries. Two energy use efficiency measures are proposed, one based on the renewability ratio (RR) and the other on the economic-output-induced renewability ratio (E-RR). 3.3.1. Mathematical Framework of DEA

DEA general input-oriented model notations are proposed by Cooper et al. (1978) and Charnes et al (1978) and are indicated as follows: ∑#

 ! "


 = ∑*$% & (

(7)

'$% ' ')

subject to ∑0 ,1 +, -,. = 1

,

(8)

where j is number of service units, 2345 is service unit number j, θ is the efficiency rating of the service

unit being evaluated by DEA, 67 is the amount of output r used by service unit j, -,7 is the amount of input (i), used by service unit j, i is the number of inputs used by 2345 , r is the number of outputs

generated by 2345 , 8 is the coefficient or weight assigned by DEA to output r, and +, is the coefficient or weight assigned by DEA to input i. Equations 9-12 describe the representation of different DMUs. 234 =

% !%% 9: !%: 9⋯9 ! %

(9)

 !

(101)

&% (%% 9&: (:% 9⋯9&* (*%

9: !:: 9⋯9 ! : % %: : (:: 9⋯9 &* (*%

234< = & %( %:9& …  !

9 !

9⋯9 ! " * (*"

234. = & %( %"9& :( :"9⋯9& % %"

: :"

(11)

… 2347 =

% !%) 9: !:) 9⋯9 ! )

&% (%) 9&: (:) 9⋯9&* (*)

(12)

17

where 8 , … , 8 ≥ 0 @A + , … +0 ≥ 0. 3.3.2. Energy Use Efficiency Measure 1: Renewability Ratio (RR) The renewability ratio (RR) has been used in a number of studies. For instance, Kucukvar and Tatari (2011) defined the RR as “the ratio of renewable resources used to the total ecological resources consumed and [it] offers a valuable indication of renewable resources depletion.” Moreover, Coskun et al. (2010), in another ecologically-focused sustainability assessment study, proposed an exergetic and energetic renewability ratio for assessing geothermal district heating systems. Other examples include Balta et al. (2010), Coskun et al. (2011, 2012), Hepbasli (2011), and Egilmez et al. (2015). This study focuses on the life-cycle inventory (LCI) exclusively in the context of aggregated renewable energy use and aggregated nonrenewable energy use. Thus, renewability ratio is defined as the   ) to the total nonrenewable energy use (C.CBC ) of a ratio of total renewable energy use (BC ,, ,,

manufacturing industry i in year t. It is important to note that the term total energy use indicates the total amount of energy used in a specific industry across the globe, which includes domestic and global energy use for that specific industry, whether it be renewable or nonrenewable. The data for energy use (both renewable and nonrenewable) were obtained from the MRIO experiments, where the energy-use impacts of the U.S. manufacturing industries were traced at the onsite production, domestic supply chain, and global supply chain levels. Equation 13 depicts the RR measure for industry i, where i = 1… 35, c is energy carrier (among 9 renewable and 7 nonrenewable energy carriers—see Table 2), in year t (from 1995 to 2014). In equation 13, each industry (i) is considered to be a DMU. Furthermore, to benchmark the countries, the industries’ energy use was aggregated, and each country was treated as a DMU: DD,, =

9  ∑41 H1 ∑F1 BC ,,

16  ∑41 H1 ∑F18 C.CBC ,,

KH i = 1 … 35,

O = 1,2, … 20 . (13

The aims of the input-oriented DEA approach were to conduct a RR efficiency analysis of 35 industries (35 DMUs were studied comparatively) and then of the 41 countries (41 DMUs were studied

18

comparatively). The experiments were conducted roughly every five years – in 1995, 2000, 2005, 2010 and 2014 – to identify trends in RR-efficiency change. To conduct the RR efficiency assessment, first the renewable and nonrenewable energy use impacts of U.S. manufacturing activities were obtained from the experimental results of the MRIO-1995, MRIO-2000, MRIO-2005, MRIO-2010, and MRIO-2014 models. Next, the results for renewable and nonrenewable energy use were aggregated by industry, and lastly, 35 DEA models were built, and experiments conducted for all the industries, to quantity the RR efficiency scores. 3.3.3. Energy Use Efficiency Measure 2: Economic-Output-Induced Renewability Ratio (E-RR) For this measure, the efficiency assessment focus is termed the economic-output-induced renewability ratio (E-RR). In the first measure, only energy use impacts were considered, and the focus was on the ratio of total renewable energy use to total nonrenewable energy use. For this measure, economic output was added to make a more comprehensive efficiency assessment (see equation 14). The mean normalization approach was used to normalize the LCI results (the data for both renewable and nonrenewable energy and economic output) to ensure that the differences in units of measurement (energy use: TJs, economic output: $s) did not skew the results of the experiments (Egilmez et al., 2013; VázquezRowe et al., 2012; Park et al., 2016). DD,, =

 S  U R∑U 1 ∑T1 BC ,, V + W∑1
,, X U  ∑U 1 ∑T1Y C.CBC ,,

KH i = 1 … 35,

O = 1,2, … 20 (14

4. Results 4.1. MRIO Results 4.1.1. Economic Output The domestic (onsite supply chains) and global economic output shares (%) of the U.S. manufacturing industries are provided in Figure 2. According to the results, the total economic output of the industries increased during the study period, such that the domestic (U.S. manufacturing activities and

19

supporting supply chain activities in the U.S.) and global supply chains had shares of 80% to 90% and 10% to 20% of total economic output, respectively. Total Output of Global

Total Output of USA

$7.00 $6.56 $6.31 $6.01 $6.00

$5.65

$5.41 $5.49

2009

2010

1998

2008

1997

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

[CELLRANGE], [VALUE]

$1.00

[CELLRANGE], [VALUE]

$2.00

[CELLRANGE], [VALUE]

$3.00

$3.86

$3.41 [CELLRANGE], [VALUE]

$3.28

$3.97

$3.65 $3.69

[CELLRANGE], [VALUE]

$3.91 [CELLRANGE], [VALUE]

$4.00

$4.02

2007

$4.36 $4.14

[CELLRANGE], [VALUE]

$4.59 $4.64

[CELLRANGE], [VALUE]

$4.80

$5.00

[CELLRANGE], [VALUE]

$5.11

2014

2013

2012

2011

Axis Title

2006

2005

2004

2003

2002

2001

2000

1999

1996

1995

$-

Figure 2. U.S. Domestic and global economic output shares ($) of U.S. manufacturing industries. The results of the economic analysis (Figure 2) show that the U.S. manufacturing industries’ output was the lowest in 1995, as opposed to its having the highest share of total global output, at 89%. Furthermore, the share of domestic total output decreased gradually from 1995 to 2014, while direct economic output is increasing. This indicates that its global supply chain dependency is financially increasing. The domestic economic output average was seen to be 3.93 million dollars. It reached its maximum output level in 2014 at 5.36 million dollars and its minimum level in 1995 at 2.91 million dollars. The domestic and total output levels show that similar production output behaviors and their

20

output levels gradually increased until 2008. In that year, the U.S. suffered an economic crisis, which had an adverse effect on the domestic output level and resulted in a one-million-dollar decline between 2008 and 2009. The bar chart shows that domestic and global output levels started to rise again in 2010, which also indicates a recovery trend in the manufacturing industry between 2010 and 2014. 4.1.2. Energy Use Figure 3 presents the nonrenewable energy usage of U.S. manufacturing industries associated with their domestic and global supply chain linkages. The time series graph covers the years 1995 through 2014.

3.9E+07

3.4E+07

2.9E+07

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

2.4E+07

Figure 3. Nonrenewable energy use (in TJs). According to the results in Figure 3, the U.S. consumed the most nonrenewable energy in year 2014 (37.9 million TJs) and the lowest usage in 1995 (24.6 million TJs). In addition, the average consumption for nonrenewable energy source was recorded as 28.4 million TJs/yr. Obviously, the economic crisis in 2008 affected domestic and global output levels negatively, and the U.S. manufacturing industries produced and used less energy, both renewable and nonrenewable, compared to other years. It was observed that there is a sharp increase in nonrenewable energy use between 2009 and

21

2012. Majority of the factors caused this significant increase in the nonrenewable use were attributed to 2008 financial crisis. The global oil price went down from $164.64 to $50.81 just in the second half of 2008 (“Crude Oil Prices - 70 Year Historical Chart,” 2019).The market crash significantly affected the global energy markets, and energy policies. In addition, the financing of renewable energy projects worldwide was also affected. Clean energy investments were down by 17% on the first and by 23% on the second half of 2007. The 2009 started with the lowest investment in clean energy by $13.3 billion since the first quarter of 2006 (Fritz-Morgenthal et al., 2009). It was evident that renewable energy adoption and investment is highly sensitive to the financial crisis. Figure 4 shows the renewable energy use of the U.S. manufacturing industries associated with their domestic and global supply chain linkages. The time series covers the years from 1995 to 2014. During the financial crisis and following few years, an increasing trend could be observed in renewable energy use, which is mostly attributed to the clean energy support and funding policies initiated with the American Recovery and Reinvestment Act of 2009 (ARRA) (NTIA, 2009).

22

2.0E+06

1.5E+06

1.0E+06

5.0E+05

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

0.0E+00

Figure 4. Renewable energy use (TJs). According to the results in Figure 4, the U.S. manufacturing industries consumed the most renewable energy in year 2001 (1.5 million TJs), and the lowest usage is observed in 2010 (1.1 million TJs). In addition, the average consumption for renewable energy source was recorded as 1.3 million TJs/yr. Renewable energy use shows a fluctuating trend, but the trend line indicates non-increasing (stagnant) behavior (see Figure 4). From both Figure 3 and 4, the steep increase in usage of both renewable and nonrenewable energy between 2009 and 2012 represents the recovery of the US manufacturing economic impacts. It is important to note that final demand and government policies play a significant role in whether companies use renewable energy or not. Some policies relieve taxes on companies when they use more renewable energy in a given specific year (Shafiee & Topal, 2009). However, some mismatches can be observed between renewable energy use and the total output level graphs, because most of the time, the U.S. manufacturing industries’ first choice is nonrenewable energy,

23

and the destiny of renewable energy usage depends more on effective, strategic, government-level policy making. Figure 5 shows the top five energy carriers impacted by the U.S. manufacturing industries’ economic output throughout the study period. This graph is crucial for revealing the energy use of the U.S. manufacturing industries and their global supply chains by carrier type and percentage share.

BCOAL BIOGAS HCOAL NATGAS

0.48% 3.71% 12.49% 16.54%

CRUDE

65.47%

Figure 5. Use of the top five energy carriers’ from 1995 to 2014. Crude was the energy carrier used most by the U.S. manufacturing industries and their domestic and global supply chains in the 19-year period, due to its domestic production and supply chain linkages with other countries. It is in top place, with a share of 65.47%. Unfortunately, only one renewable energy carrier was found to be in the top five most-used energy carriers. Biogas is in fourth place as a form of renewable energy used by domestic manufacturing industries, with a percentage of 3.71%. Figure 6 shows the top five most-used energy carriers in 2014. It enables a comparison to be made with Figure 5 (which shows the top five energy carriers used by by U.S. manufacturing industries from 1995 to 2014) to see if any changes happened in the U.S. manufacturing industries’ preferences for specific energy carriers. According to Figure 6, crude was still the locomotive of the U.S. manufacturing industries, at 72.08%. NatGas and HCoal were consumed mostly after crude, at 16.10% and 7.57%, respectively. Furthermore, hydro power placed fifth (0.30%) as a renewable energy source after BioGas

24

(2.77%). Thus, the top three nonrenewable energy carriers maintained their place, and in 2014 hydropower joined the top five most-used energy carrier list as a renewable energy source.

HYDRO BIOGAS HCOAL

0.30% 2.77% 7.57%

NATGAS

16.10%

CRUDE

72.08%

Figure 6. Top five energy carriers used by U.S. manufacturing industries in 2014.

4.2. Eco-Efficiency Results The RR and E-RR eco-efficiency results are presented separately, in two sections.

4.2.1. Renewability Ratio (RR) Results The changing ranking trends from the RR analysis of the industries are shown in Table 6 for the years 1995, 2000, 2005, 2010 and 2014. The RR scores are depicted with the symbols ✓, !, and x to make the interpretation of the findings easier for the reader. The check mark (✓) indicates an eco-efficiency score greater than 0.65; the exclamation mark (!) indicates that the eco-efficiency score is between 0.65 and 0.35; and the cross (x) indicates a very low eco-efficiency score lower than 0.35. Table 5 presents the descriptive statistics for the RR scores for the study period. In the RR results, it is evident that, on the average, a manufacturing industry’s RR score ranged between 0.3 and 0.4 over the course of study period, which is quite low. The good side is that there is an encouraging positive growth observed between 1995 and 2014 in the mean RR scores.

25

Table 5. Descriptive statistics for the RR scores

Average Std.Dev. Min. Max. Range

1995 0.30 0.38 0.00 1.00 1.00

2000 0.16 0.33 0.00 1.00 1.00

2005 0.30 0.39 0.00 1.00 1.00

2010 0.40 0.44 0.00 1.00 1.00

2014 0.40 0.44 0.00 1.00 1.00

The results shown in Table 6 indicate that the top five most efficient industries were Agriculture, Hunting, Forestry and Fishing (AHFF: mean RR score 0.83), Chemicals and Chemical Products (CCP: mean RR score 0.80), Manufacturing, Nec.; Recycling (MNR: mean RR score 0.72), Textiles and Textile Products (TTP: mean RR score 0.60), and Leather, Leather and Footwear (LLF: mean RR score 0.56). The industries with the lowest efficiency scores were Other Non-Metallic Mineral (ONM), Transport Equipment (TE), and Food, Beverages and Tobacco (FBT), with mean RR scores of 0.03, and Basic Metals and Fabricated Metal (BMFM) and Coke, Refined Petroleum and Nuclear Fuel (CRPNF), with mean RR scores of 0.01 and 0.00, respectively.

26

Table 6. Results of Measure 1: Renewability ratio (RR) analysis (ranked on the basis of the mean RR scores) DMU No. 1 9 16 4 5 6 14 10 2 7 13 11 15 3 12 8

DMU Name Agriculture, Hunting, Forestry and Fishing Chemicals and Chemical Products Manufacturing, Nec; Recycling Textiles and Textile Products Leather, Leather and Footwear Wood and Products of Wood and Cork Electrical and Optical Equipment Rubber and Plastics Mining and Quarrying Pulp, Paper, Paper , Printing and Publishing Machinery, Nec Other Non-Metallic Mineral Transport Equipment Food, Beverages and Tobacco Basic Metals and Fabricated Metal Coke, Refined Petroleum and Nuclear Fuel

1995 1.00 1.00 0.56 0.07 0.08 0.32 1.00 0.18 0.01 0.29 0.09 0.04 0.07 0.02 0.04 0.00

2000 1.00 0.00 0.04 0.00 0.02 0.09 0.00 0.06 1.00 0.28 0.00 0.00 0.00 0.08 0.01 0.00

2005 0.17 1.00 1.00 1.00 0.71 0.24 0.17 0.30 0.00 0.08 0.02 0.04 0.02 0.01 0.00 0.00

2010 1.00 1.00 1.00 0.96 1.00 0.55 0.20 0.51 0.00 0.09 0.03 0.04 0.02 0.02 0.01 0.00

2014 1.00 1.00 1.00 0.96 1.00 0.55 0.20 0.51 0.00 0.09 0.03 0.04 0.02 0.02 0.01 0.00

Min. 0.17 0.00 0.04 0.00 0.02 0.09 0.00 0.06 0.00 0.08 0.00 0.00 0.00 0.01 0.00 0.00

Mean Median Max. St.Dev. 0.83 1.00 1.00 0.37 0.80 1.00 1.00 0.45 0.72 1.00 1.00 0.42 0.60 0.96 1.00 0.51 0.56 0.71 1.00 0.48 0.35 0.32 0.55 0.20 0.32 0.20 1.00 0.39 0.31 0.30 0.51 0.20 0.20 0.00 1.00 0.45 0.17 0.09 0.29 0.11 0.03 0.03 0.09 0.03 0.03 0.04 0.04 0.02 0.03 0.02 0.07 0.02 0.03 0.02 0.08 0.03 0.01 0.01 0.04 0.02 0.00 0.00 0.00 0.00

: RR≥0.65, : 0.35
27

4.2.2. Economic-Output-Induced Renewability Ratio (E-RR) Results The descriptive statistics by study year are shown in Table 7 for the years 1995, 2000, 2005, 2010, and 2014. On the other hand, industry-specific results and their descriptive statistics are provided in Table 8. Table 7 indicates positive growth in the industries’ average E-RR scores, with a low E-RR score as a single exception in 2000. It is questionable whether this can be interpreted as a positive result, since the maximum E-RR measure exceeded 0.5 in 2014. In this context, it is important to note that E-RR measures consider economic output as a positive output in addition to renewable energy, while keeping nonrenewable energy use as the input, which is deemed to be minimized. Table 75. Mean and standard deviations for the E-RR analysis 1995 0.38 0.40 0.00 1.00 1.00

Average Std.Dev. Min. Max. Range

2000 0.29 0.43 0.00 1.00 1.00

2005 0.53 0.45 0.00 1.00 1.00

2010 0.52 0.47 0.00 1.00 1.00

2014 0.52 0.47 0.00 1.00 1.00

According to the industry-specific E-RR results given in Table 8, the top five most efficient industries were found to be Electrical and Optical Equipment (EOE: mean E-RR score 1.00), Agriculture, Hunting, Forestry and Fishing (AHFF: mean E-RR score 0.96), Food, Beverages and Tobacco (FBT: mean E-RR score 0.85), Chemicals and Chemical Products (CCP: mean E-RR score 0.80), and Manufacturing, Nec.; Recycling (MNR: mean E-RR score 0.72). On the other hand, the industries with the lowest five E-RR scores are observed to be Pulp, Paper, Paper, Printing and Publishing (PPPP: mean E-RR score 0.17), Machinery, Nec (MN: mean E-RR score 0.10), Other Non-Metallic Mineral (ONM: mean E-RR score 0.03), Basic Metals and Fabricated Metal (BMFM: mean E-RR score 0.02), and Coke, Refined Petroleum and

Nuclear

Fuel

(CRPNF:

mean

E-RR

score

0.00).

28

Table 8. Results of the E-RR analysis (ranked on the basis of the mean E-RR scores) DMU No. DMU Name 14 Electrical and Optical Equipment 1 Agriculture, Hunting, Forestry and Fishing 3 Food, Beverages and Tobacco 9 Chemicals and Chemical Products 16 Manufacturing, Nec; Recycling 4 Textiles and Textile Products 5 Leather, Leather and Footwear 6 Wood and Products of Wood and Cork 10 Rubber and Plastics 15 Transport Equipment 2 Mining and Quarrying 7 Pulp, Paper, Paper , Printing and Publishing 13 Machinery, Nec 11 Other Non-Metallic Mineral 12 Basic Metals and Fabricated Metal 8 Coke, Refined Petroleum and Nuclear Fuel

1995 1.00 1.00 0.26 1.00 0.56 0.07 0.08 0.32 0.18 1.00 0.01 0.29 0.22 0.04 0.04 0.00

2000 1.00 1.00 1.00 0.01 0.04 0.00 0.02 0.21 0.06 0.00 1.00 0.28 0.00 0.00 0.01 0.00

2005 1.00 0.81 1.00 1.00 1.00 1.00 0.71 1.00 0.52 0.09 0.02 0.10 0.19 0.04 0.02 0.00

2010 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.58 0.51 0.02 0.03 0.09 0.05 0.04 0.01 0.00

2014 Min. 1.00 1.00 1.00 0.81 1.00 0.26 1.00 0.01 1.00 0.04 1.00 0.00 1.00 0.02 0.58 0.21 0.51 0.06 0.02 0.00 0.03 0.01 0.09 0.09 0.05 0.00 0.04 0.00 0.01 0.01 0.00 0.00

Mean 1.00 0.96 0.85 0.80 0.72 0.62 0.56 0.54 0.36 0.23 0.22 0.17 0.10 0.03 0.02 0.00

Median Max. 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.71 1.00 0.58 1.00 0.51 0.52 0.02 1.00 0.03 1.00 0.10 0.29 0.05 0.22 0.04 0.04 0.01 0.04 0.00 0.00

St.Dev. 0.00 0.09 0.33 0.44 0.42 0.53 0.48 0.31 0.22 0.43 0.44 0.11 0.10 0.02 0.01 0.00

: E-RR≥0.65, : 0.35
29

4.2.3. Comparison of RR and E-RR Measures This section introduces the graphical and statistical comparison of the two measures, RR and E-RR, respectively. The graphical comparisons are shown in Figure 7, which features a bar chart of the industries’ mean RR and mean E-RR scores for the study period. The RR measure do not take into account the economic output but focus only on the ratio of renewable energy use to non-renewable energy use. On the other hand, E-RR considers the economic output as a positive output variable. The RR scores were found to be lower than the E-RR scores, because the RR measure focuses solely on environmental sustainability aspects, while the E-RR scores consider both environmental and economic sustainability domains. The RR measure is obviously a more conservative measure than the E-RR. However, both measures do indicate a quite similar trend in terms of the most- and least-efficient industries, as the rankings of the industries do not shift between the two measures.

Manufacturing, Nec; Recycling Transport Equipment Electrical and Optical Equipment Machinery, Nec Basic Metals and Fabricated Metal Other Non-Metallic Mineral Rubber and Plastics Chemicals and Chemical Products Coke, Refined Petroleum and Nuclear Fuel Pulp, Paper, Paper , Printing and Publishing Wood and Products of Wood and Cork Leather, Leather and Footwear Textiles and Textile Products Food, Beverages and Tobacco Mining and Quarrying Agriculture, Hunting, Forestry and Fishing

0.00

0.20

0.40

Mean ERR

0.60

0.80

1.00

1.20

1.40

Mean RR

Figure 7. Comparison of the mean RR and E-RR scores by industry.

30

In addition, a statistical comparison of the RR and E-RR results for the study period was conducted by using a one-way ANOVA test, whose results are shown in Table 9. The study periods included in the ANOVA tests were 1995, 2000, 2005, 2010, and 2014. The objective was to see if there is significant statistical difference in the RR scores among the years and in the ERR scores among the years. The statistical comparison results do not indicate a significant difference, as all the p-values were found to be greater than 0.05 (df = 5 study periods x 16 industries minus 1; namely, 80 -1 = 79). This could be interpreted that no substantial improvement was achieved in both the RR and E-RR scores of the industries over the 20-year study period. This is also in parallel with the findings of the MRIO analysis, which indicate that a great reliance on nonrenewable energy sources still dominates the overall energy use of the manufacturing industries and their domestic and global supply chains.

Table 9. Results of the ANOVA tests ANOVA by RR Source of Variation SS Between Groups 0.626597 Within Groups 12.00736 Total

15.40773

MS

F

P-value

F crit

4 0.156649 0.978457 0.424522 2.493696 75 0.160098

12.63396

ANOVA by ERR Source of Variation SS Between Groups 0.752216 Within Groups 14.65552 Total

df

79

df

MS

F

P-value

F crit

4 0.188054 0.962372 0.433257 2.493696 75 0.195407 79

5. Conclusion: Limitations, and Future Work This paper introduces an integrated cradle-to-gate MRIO and DEA approach to investigate the global renewable and nonrenewable energy dependency of U.S. manufacturing industries between 1995 and

31

2014. Two energy-use-focused eco-efficiency measures are proposed: RR and E-RR. The results indicate that there is still substantial reliance on nonrenewable energy sources in the activities of U.S. manufacturing industries and their domestic and global supply chains. The trend in the use of renewable energy was found to be stagnant, while nonrenewable energy use has been increasing in parallel with a growth in the economic output of U.S. manufacturing. These findings were supported by the two newly proposed energy-focused eco-efficiency measures (RR and E-RR). Both of the eco-efficiency performance indicators showed that the manufacturing industries’ relative ratio of renewable to nonrenewable energy use is not increasing. Even considering economic output in the E-RR measure did not reveal a significant increase in renewable energy use performance. This finding was also supported by the statistical tests, which yielded no statistical differences among the comparison of study periods from 1995 to 2014 in terms of RR and E-RR measures. Manufacturing industries are essential for sustainable economic growth. However, their reliance on nonrenewable energy sources substantially affects environmental sustainability and the overall sustainable development goals of the nations in the world, since U.S. manufacturing is one of the most dominant influences on the global economy. The literature indicates that the intensity of the U.S. manufacturing industries’ energy use (both renewable and nonrenewable) has declined over the last two decades, due to technological advancements and enhanced competitiveness through effective costreduction strategies, which have resulted in less energy-intense production (Egilmez et al., 2017). However, even though the energy intensity is decreasing, total energy use is increasing along with the total economic output. If we consider two facts: (1) that the amount of GHG emissions is significantly correlated with energy use and shows a non-decreasing trend (Egilmez et al., 2019) and (2) that the U.S. manufacturing industries’ total energy use is increasing, and its share of renewable energy use has stagnated at around 5%-10% of the total, this growth is not well aligned with the sustainable development goals (SDGs) accepted and set by the United Nations in 2015. If sufficient action is not taken, the manufacturing activities of the U.S. will continue to substantially contribute to GHG emissions, and alarm bells will be ringing worldwide, since the current U.S. administration has already withdrawn the support

32

of the U.S. for the UN 2030 Sustainable Development Agenda (17 SDGs), and since U.S. economic production, as well as consumption, plays an instrumental role in global production and consumption, which has immediate social, economic, and environmental implications. These actions call for betterstructured and better-aimed economic stimulation and incentive plans to be made available to manufacturing enterprises that adopt circular design and production techniques and technologies. This would not only substantially decrease the environmental consequences of production activities overall, but also make U.S. manufacturing more competitive, since a circular economy aims to reduce productionrelated material and other waste to zero, ensures a substantial use of clean renewable energy, and paves the way for a transition in a country’s economic expansion to a circular rather than a linear economy (Europe’s circular-economy opportunity, 2015; Macarthur, 2015). For the U.S., which has an immense capacity to endlessly landfill any waste (supposedly) and which is among the top energy-exporter countries, it may look quite counterintuitive to adopt circular economic growth principles and strategies across all industries. However, we cannot and should not underestimate the competitive edge that these strategies are already bringing by way of efficiency to the labor, material, and energy costs of production activities in Europe and Asian markets, as indicated in a recent U.S. manufacturing health report (Levinson, 2018). This study has the following limitations, which could be further investigated and left as future work. The study focused only on manufacturing industries in the U.S. with their global supply chain linkages. This scope can be extended to include or focus on other industries, such as agriculture, service, construction, etc., and comparative analyses could be made. The dependence of other countries’ manufacturing industries on renewable versus nonrenewable energy could be studied and compared with the U.S. In terms of energy carriers, this study focused its eco-efficiency analysis on aggregated renewable and nonrenewable energy carriers. This could be further extended by conducting a comparative analysis of each of the renewable and nonrenewable energy carriers. In terms of the methods, in addition to DEA, stochastic frontier analysis (SFA) could be used to quantify eco-efficiency, and a comparative

33

analysis of DEA and SFA-based eco-efficiency could provide further information on the methodological implications of the results.

References Abbod, K. (2016). Multi-region Input-Output-based Carbon and Energy Footprint Analysis of U.S.

Manufacturing. University of New Haven. Ali, Y., Pretaroli, R., Socci, C., & Severini, F. (2018). Carbon and water footprint accounts of Italy: A Multi-Region Input-Output approach. Renewable and Sustainable Energy Reviews, 81(May), 1813– 1824. https://doi.org/10.1016/j.rser.2017.05.277 Andrew, R., Peters, G. P., & Lennox, J. (2009). Approximation and regional aggregation in multiregional input-output analysis for national carbon footprint accounting. Economic Systems Research,

21(3), 311–335. https://doi.org/10.1080/09535310903541751 Balta, M. T., Dincer, I., & Hepbasli, A. (2010). Potential methods for geothermal-based hydrogen production. International Journal of Hydrogen Energy, 35(10), 4949–4961. https://doi.org/10.1016/j.ijhydene.2009.09.040 BEA - Gross Output of U.S. Manufacturing. (2018). Retrieved November 23, 2018, from https://apps.bea.gov/iTable/iTable.cfm?ReqID=51&step=1#reqid=51&step=2&isuri=1 Charnes, A., Cooper, W. W., & Rhodes, E. (1978). Measuring the efficiency of decision-making units.

European Journal of Operational Research, 3(4), 339–338. Chen, B., Li, J. S., Wu, X. F., Han, M. Y., Zeng, L., Li, Z., & Chen, G. Q. (2018). Global energy flows embodied in international trade: A combination of environmentally extended input–output analysis and complex network analysis. Applied Energy, 210(July 2017), 98–107. https://doi.org/10.1016/j.apenergy.2017.10.113 Chenel, S. (2014). ScienceDirect Eco-efficiency analysis of Spanish WWTPs using the LCA þ DEA method, 8. https://doi.org/10.1016/j.watres.2014.10.040 Chien, T., & Hu, J. L. (2007). Renewable energy and macroeconomic efficiency of OECD and non-

34

OECD economies. Energy Policy, 35(7), 3606–3615. https://doi.org/10.1016/j.enpol.2006.12.033 Cooper, W. W., Seiford, L. M., & Zhu, J. (n.d.). Chapter 1, 1–39. Coskun, C., Oktay, Z., & Dincer, I. (2010). Investigation of some renewable energy and exergy parameters for two Geothermal District Heating Systems. International Journal of Exergy, 8(1), 1. https://doi.org/10.1504/ijex.2011.037211 Coskun, C., Oktay, Z., & Dincer, I. (2011). Performance evaluations of a geothermal power plant.

Applied Thermal Engineering, 31(17–18), 4074–4082. https://doi.org/10.1016/j.applthermaleng.2011.08.013 Coskun, C., Oktay, Z., & Dincer, I. (2012). Thermodynamic analyses and case studies of geothermal based multi-generation systems. Journal of Cleaner Production, 32, 71–80. https://doi.org/10.1016/j.jclepro.2012.03.004 Crude Oil Prices - 70 Year Historical Chart. (2019). Retrieved January 11, 2019, from https://www.macrotrends.net/1369/crude-oil-price-history-chart Database, W. I. O. (n.d.). No Title. WIOD. Dietzenbacher, E., Los, B., Stehrer, R., Timmer, M., & de Vries, G. (2013). the Construction of World Input-Output Tables in the Wiod Project. Economic Systems Research, 25(1), 71–98. https://doi.org/10.1080/09535314.2012.761180 Egilmez, G., Bhutta, K., Erenay, B., Park, Y. S., & Gedik, R. (2017). Carbon footprint stock analysis of US manufacturing: A time series input-output LCA. Industrial Management and Data Systems,

117(5), 853–872. https://doi.org/10.1108/IMDS-06-2016-0253 Egilmez, G., Bhutta, K., Erenay, B., Park, Y. S., & Gedik, R. (2017). Carbon footprint stock analysis of US manufacturing: A time series input-output LCA. Industrial Management and Data Systems,

117(5). https://doi.org/10.1108/IMDS-06-2016-0253 Egilmez, G., Kucukvar, M., & Park, Y. S. (2015). Supply chain-linked sustainability assessment of the US manufacturing: An ecosystem perspective. Sustainable Production and Consumption, 5(October 2015), 65–81. https://doi.org/10.1016/j.spc.2015.10.001

35

Egilmez, G., Kucukvar, M., & Park, Y. S. (2016a). Mode-specific eco-efficiency analysis of the freight transportation in the USA: an integrated life cycle assessment and linear programming approach.

World Review of Intermodal Transportation Research, 6(1), 16. https://doi.org/10.1504/writr.2016.078137 Egilmez, G., Kucukvar, M., & Park, Y. S. (2016b). Supply chain-linked sustainability assessment of the US manufacturing: An ecosystem perspective. Sustainable Production and Consumption, 5, 65–81. Egilmez, G., Kucukvar, M., & Tatari, O. (2013). Sustainability assessment of U.S. manufacturing sectors: An economic input output-based frontier approach. Journal of Cleaner Production, 53, 91–102. https://doi.org/10.1016/j.jclepro.2013.03.037 Egilmez, G., Kucukvar, M., Tatari, O., & Bhutta, M. K. S. (2014). Supply chain sustainability assessment of the U.S. food manufacturing sectors: A life cycle-based frontier approach. Resources,

Conservation and Recycling, 82, 8–20. https://doi.org/10.1016/j.resconrec.2013.10.008 Egilmez, G., Oztanriseven, F., & Gedik, R. (2019). Energy, Climate, Water Nexus: A Multi-Region Input Output Analysis of U.S. Manufacturing. Engineering Management Journal, (Revised). EPA. (2019). Inventory of U.S. greenhouse gas emissions and sinks. 1990-2011. Retrieved from https://www.epa.gov/sites/production/files/2019-04/documents/us-ghg-inventory-2019-maintext.pdf

Europe’s circular-economy opportunity. (2015). Retrieved from https://www.mckinsey.com/businessfunctions/sustainability/our-insights/europes-circular-economy-opportunity Fritz-Morgenthal, S., Chris, G., Carola, M., Marija, M., & Virginia, S. (2009). The global financial crisis

and its impact on renewable energy finance. Frankfurt, Germany. Gokhan Egilmez, N. Muhammad Aslaam Mohamed Abdul Ghani, R. G. (2017). Industrial Management & Data Systems Article information : Layer Analysis of CO 2 Sources in the U . S . Economic Supply Chains : An Input Output LCA Study Abstract. Hamilton, T. G. A., & Kelly, S. (2017). Low carbon energy scenarios for sub-Saharan Africa: An inputoutput analysis on the effects of universal energy access and economic growth. Energy Policy,

36

105(July 2016), 303–319. https://doi.org/10.1016/j.enpol.2017.02.012 Hepbasli, A. (2011). A comparative investigation of various greenhouse heating options using exergy analysis method. Applied Energy, 88(12), 4411–4423. https://doi.org/10.1016/j.apenergy.2011.05.022 Hong, J., Shen, G. Q., Guo, S., Xue, F., & Zheng, W. (2016). Energy use embodied in China’s construction industry: A multi-regional input-output analysis. Renewable and Sustainable Energy

Reviews, 53, 1303–1312. https://doi.org/10.1016/j.rser.2015.09.068 Honma, S., & Hu, J. L. (2014). A panel data parametric frontier technique for measuring total-factor energy efficiency: An application to Japanese regions. Energy, 78, 732–739. https://doi.org/10.1016/j.energy.2014.10.066 Hosseinzadeh-Bandbafha, H., Safarzadeh, D., Ahmadi, E., Nabavi-Pelesaraei, A., & HosseinzadehBandbafha, E. (2017). Applying data envelopment analysis to evaluation of energy efficiency and decreasing of greenhouse gas emissions of fattening farms. Energy, 120, 652–662. https://doi.org/10.1016/j.energy.2016.11.117 Huppes, G., & Ishikawa, M. (2005). Eco-efficiency and Its xsTerminology. Journal of Industrial Ecology,

9(4), 43–46. https://doi.org/10.1162/108819805775247891 Iribarren, D., & Vázquez-Rowe, I. (2013). Is Labor a Suitable Input in LCA + DEA Studies? Insights on the Combined Use of Economic, Environmental and Social Parameters. Social Sciences, 2(3), 114– 130. https://doi.org/10.3390/socsci2030114 Kouchaki-Penchah, H., Nabavi-Pelesaraei, A., O’Dwyer, J., & Sharifi, M. (2017). Environmental management of tea production using joint of life cycle assessment and data envelopment analysis approaches. Environmental Progress and Sustainable Energy, 36(4), 1116–1122. https://doi.org/10.1002/ep.12550 Kucukvar, M., Cansev, B., Egilmez, G., Onat, N. C., & Samadi, H. (2016). Energy-climatemanufacturing nexus: New insights from the regional and global supply chains of manufacturing industries. Applied Energy, 184, 889–904. https://doi.org/10.1016/j.apenergy.2016.03.068

37

Kucukvar, M., Onat, N. C., Abdella, G. M., & Tatari, O. (2019). Assessing regional and global environmental footprints and value added of the largest food producers in the world. Resources,

Conservation and Recycling, 187–197. https://doi.org/10.1016/j.resconrec.2019.01.048 Kucukvar, M., & Tatari, O. (2011). A comprehensive life cycle analysis of cofiring algae in a coal power plant as a solution for achieving sustainable energy. Energy, 36(11), 6352–6357. https://doi.org/10.1016/j.energy.2011.09.039 Lenzen, M., Wood, R., & Wiedmann, T. (2010). Uncertainty analysis for multi-region input - output models - a case study of the UK’S carbon footprint. Economic Systems Research, 22(1), 43–63. https://doi.org/10.1080/09535311003661226 Leontief, W. (1970). Environmental Repercussions and the Economic Structure: An Input-Output Approach. REStat, 52(3), 262–271. Retrieved from http://www.jstor.org/discover/10.2307/1926294?uid=3739840&uid=2129&uid=2&uid=70&uid=4& uid=3739256&sid=21101262763267 Levinson, M. (2018). U.S. Manufacturing in International Perspective. https://doi.org/10.17226/13497 Macarthur, E. (2015). Growth within: a circular economy vision for a competitive europe. Ellen

MacArthur Foundation. Miketa, A., & Mulder, P. (2005). Energy productivity across developed and developing countries in 10 manufacturing sectors: Patterns of growth and convergence. Energy Economics, 27(3), 429–453. https://doi.org/10.1016/j.eneco.2005.01.004 Miller, R. E., & Blair, P. D. (n.d.). Input – Output Analysis. Mukherjee, K. (2008). Energy use efficiency in U.S. manufacturing: A nonparametric analysis. Energy

Economics, 30(1), 76–96. https://doi.org/10.1016/j.eneco.2006.11.004 Nabavi-Pelesaraei, A., Abdi, R., Rafiee, S., & Mobtaker, H. G. (2014). Optimization of energy required and greenhouse gas emissions analysis for orange producers using data envelopment analysis approach. Journal of Cleaner Production, 65, 311–317. https://doi.org/10.1016/j.jclepro.2013.08.019

38

Nabavi-Pelesaraei, A., Abdi, R., Rafiee, S., & Taromi, K. (2014). Applying data envelopment analysis approach to improve energy efficiency and reduce greenhouse gas emission of rice production.

Engineering in Agriculture, Environment and Food, 7(4), 155–162. https://doi.org/10.1016/j.eaef.2014.06.001 Nakano, S., Arai, S., & Washizu, A. (2018). Development and application of an inter-regional inputoutput table for analysis of a next generation energy system. Renewable and Sustainable Energy

Reviews, 82(July), 2834–2842. https://doi.org/10.1016/j.rser.2017.10.011 NTIA. (2009). American Recovery and Reinvestment Act of 2009. Washington, DC. Retrieved from https://www.govinfo.gov/content/pkg/BILLS-111hr1enr/pdf/BILLS-111hr1enr.pdf Palmer, G. (2017). An input-output based net-energy assessment of an electricity supply industry. Energy,

141, 1504–1516. https://doi.org/10.1016/j.energy.2017.11.072 Pan, L., Liu, P., Li, Z., & Wang, Y. (2017). A dynamic input–output method for energy system modeling and analysis. Chemical Engineering Research and Design. https://doi.org/10.1016/j.cherd.2017.11.032 Park, Y. S., Egilmez, G., & Kucukvar, M. (2015). A Novel Life Cycle-based Principal Component Analysis Framework for Eco-efficiency Analysis: Case of the United States Manufacturing and Transportation Nexus. Journal of Cleaner Production, 92, 327–342. https://doi.org/10.1016/j.jclepro.2014.12.057 Park, Y. S., Egilmez, G., & Kucukvar, M. (2016). Emergy and end-point impact assessment of agricultural and food production in the United States: A supply chain-linked Ecologically-based Life Cycle Assessment. Ecological Indicators, 62, 117–137. https://doi.org/10.1016/j.ecolind.2015.11.045 Phylipsen, G. J. M., Blok, K., & Worrell, E. (1997). International comparisons of energy efficiencyMethodologies for the manufacturing industry. Energy Policy, 25(7–9), 715–725. https://doi.org/10.1016/S0301-4215(97)00063-3 Ramanathan, R. (2005). An analysis of energy consumption and carbon dioxide emissions in countries of

39

the Middle East and North Africa. Energy, 30, 2831–2842. https://doi.org/10.1016/j.energy.2005.01.010 Sembill, D., & Dreyer, K. (2009). Granting time—restricting time—wasting time? Subtle patterns in

vocational schooling BT - VET boost: towards a theory of professional competencies—Essays in Honor of Frank Achtenhagen (Vol. 164). https://doi.org/10.1007/978-1-4419-6151-8 Shabanzadeh-Khoshrody, M., Azadi, H., Khajooeipour, A., & Nabavi-Pelesaraei, A. (2016). Analytical investigation of the effects of dam construction on the productivity and efficiency of farmers.

Journal of Cleaner Production, 135, 549–557. https://doi.org/10.1016/j.jclepro.2016.06.145 Shafiee, S., & Topal, E. (2009). When will fossil fuel reserves be diminished? Energy Policy, 37(1), 181– 189. https://doi.org/10.1016/j.enpol.2008.08.016 Sözen, A., & Alp, I. (2009). Comparison of Turkey’s performance of greenhouse gas emissions and local/regional pollutants with EU countries. Energy Policy, 37(12), 5007–5018. https://doi.org/10.1016/j.enpol.2009.06.069 Space, S., & Development, M. (1997). Chapter 2, 30–40. https://doi.org/10.1159/000144207 Su, B., & Ang, B. W. (2014). Input-output analysis of CO2emissions embodied in trade: A multi-region model for China. Applied Energy, 114, 377–384. https://doi.org/10.1016/j.apenergy.2013.09.036 Suh, S., Lenzen, M., Treloar, G. J., Hondo, H., Horvath, A., Huppes, G., … Norris, G. (2004). System Boundary Selection in Life-Cycle Inventories Using Hybrid Approaches. Environmental Science &

Technology, 38(3), 657–664. https://doi.org/10.1021/es0263745 Sun, X., Li, J., Qiao, H., & Zhang, B. (2017). Energy implications of China’s regional development: New insights from multi-regional input-output analysis. Applied Energy, 196, 118–131. https://doi.org/10.1016/j.apenergy.2016.12.088 The World Bank. (2018). Trade % of U.S. GDP. Retrieved from https://data.worldbank.org/indicator/NE.TRD.GNFS.ZS?locations=AF-US U.S. Energy Information Administration. (2018). Nonrenewable Energy Sources. Retrieved November 24, 2018, from https://www.eia.gov/energyexplained/?page=nonrenewable_home

40

U.S. Energy Information Administration. (n.d.). Use of Energy in the United States. Retrieved November 23, 2018, from https://www.eia.gov/energyexplained/index.php?page=us_energy_use US Energy Information Administration. (2013). {E}lectric {P}ower {A}nnual 2011, (October). Vázquez-Rowe, I., Villanueva-Rey, P., Iribarren, D., Teresa Moreira, M., & Feijoo, G. (2012a). Joint life cycle assessment and data envelopment analysis of grape production for vinification in the Rías Baixas appellation (NW Spain). Journal of Cleaner Production, 27, 92–102. https://doi.org/10.1016/j.jclepro.2011.12.039 Vázquez-Rowe, I., Villanueva-Rey, P., Iribarren, D., Teresa Moreira, M., & Feijoo, G. (2012b). Joint life cycle assessment and data envelopment analysis of grape production for vinification in the Rías Baixas appellation (NW Spain). Journal of Cleaner Production, 27, 92–102. https://doi.org/10.1016/j.jclepro.2011.12.039 Wiebe, K. S., Bruckner, M., Giljum, S., & Lutz, C. (2012). Calculating energy-related CO 2 emissions embodied in international trade using a global input-output model. Economic Systems Research,

24(2), 113–139. https://doi.org/10.1080/09535314.2011.643293 Wiedmann, T. (2009). A review of recent multi-region input-output models used for consumption-based emission and resource accounting. Ecological Economics, 69(2), 211–222. https://doi.org/10.1016/j.ecolecon.2009.08.026 Wiedmann, T., Wilting, H. C., Lenzen, M., Lutter, S., & Palm, V. (2011). Quo Vadis MRIO? Methodological, data and institutional requirements for multi-region input-output analysis.

Ecological Economics, 70(11), 1937–1945. https://doi.org/10.1016/j.ecolecon.2011.06.014 Wu, X. F., & Chen, G. Q. (2017). Energy use by Chinese economy: A systems cross-scale input-output analysis. Energy Policy, 108(May), 81–90. https://doi.org/10.1016/j.enpol.2017.05.048 Xia, X. H., Chen, B., Wu, X. D., Hu, Y., Liu, D. H., & Hu, C. Y. (2017). Coal use for world economy: Provision and transfer network by multi-region input-output analysis. Journal of Cleaner

Production, 143, 125–144. https://doi.org/10.1016/j.jclepro.2016.12.142 Zhang, B., Chen, Z. M., Xia, X. H., Xu, X. Y., & Chen, Y. B. (2013). The impact of domestic trade on

41

China’s regional energy uses: A multi-regional input-output modeling. Energy Policy, 63, 1169– 1181. https://doi.org/10.1016/j.enpol.2013.08.062 Zhang, B., Qiao, H., Chen, Z. M., & Chen, B. (2016). Growth in embodied energy transfers via China’s domestic trade: Evidence from multi-regional input–output analysis. Applied Energy, 184, 1093– 1105. https://doi.org/10.1016/j.apenergy.2015.09.076 Zhang, Y., Zheng, H., Yang, Z., Su, M., Liu, G., & Li, Y. (2015). Multi-regional input-output model and ecological network analysis for regional embodied energy accounting in China. Energy Policy, 86, 651–663. https://doi.org/10.1016/j.enpol.2015.08.014

42

U.S. manufacturing industries’ time-series energy use and economic output are investigated. Global and domestic supply chain impacts were considered along with onsite production activities. Multi-region Input-Output and Data Envelopment Analysis methods were integrated. 16 renewable and nonrenewable energy carriers were considered Two energy-focused eco-efficiency indicators were proposed: RR and E-RR. A 20-year study period was chosen between 1995 and 2014 based on available data.

Authors: Bahadir Ezici (BE), Gokhan Egilmez (GE), Ridvan Gedik (RG) Contributing Author(s) Term GE Conceptualization GE, RG Methodology BE, GE Software BE Validation BE Formal analysis BE, GE Investigation Resources Data Curation BE, GE, RG Writing - Original Draft GE, RG Writing - Review & Editing BE, GE Visualization GE, RG Supervision GE Project administration Funding acquisition

Definition Ideas; formulation or evolution of overarching research goals and aims Development or design of methodology; creation of models Programming, software development; designing computer programs; implementation of the computer code and supporting algorithms; testing of existing code components Verification, whether as a part of the activity or separate, of the overall replication/ reproducibility of results/experiments and other research outputs Application of statistical, mathematical, computational, or other formal techniques to analyze or synthesize study data Conducting a research and investigation process, specifically performing the experiments, or data/evidence collection Provision of study materials, reagents, materials, patients, laboratory samples, animals, instrumentation, computing resources, or other analysis tools Management activities to annotate (produce metadata), scrub data and maintain research data (including software code, where it is necessary for interpreting the data itself) for initial use and later reuse Preparation, creation and/or presentation of the published work, specifically writing the initial draft (including substantive translation) Preparation, creation and/or presentation of the published work by those from the original research group, specifically critical review, commentary or revision – including pre-or postpublication stages Preparation, creation and/or presentation of the published work, specifically visualization/ data presentation Oversight and leadership responsibility for the research activity planning and execution, including mentorship external to the core team Management and coordination responsibility for the research activity planning and execution Acquisition of the financial support for the project leading to this publication

Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: