Karla Arias

, Maria Colmenarez

Recibido: 11/04/ 2023 y Aceptado 1/11/2023

ENERLAC. Volumen VII. Número 2. Diciembre, 2023

ISSN: 2602-8042 (impreso) / 2631-2522(digital)

Energy eciency and environmental

productivity: Analysis of Ecuadorian oil

companies

1.- Banco Interamericano de Desarrollo, kariasmarin@iadb.org, Energy Economist Consultant

0000-0003-3653-551X

2.- FLACSO. Ecuador, mgabriela3311@gmail.com

Asistente de Investigación del Centro de Estudios para el Desarrollo y Economía Aplicada

This study delves into the imperative to mitigate greenhouse gas emissions within the oil sector by

promoting energy eciency and environmental productivity. Specically, it investigates the primary

drivers inuencing eciency and productivity in private oil companies operating in Ecuador, a key South

American oil producer. The overarching research objective is to discern the factors impacting energy

eciency and productivity while considering both polluting and non-polluting aspects of productivity

variation.

Our analysis encompasses a sample of 18 Ecuadorian private oil companies, spanning the years 2012-

2020. We employ a non-parametric model and the Malmquist index to comprehensively assess energy

eciency and productivity in two distinct scenarios, accounting for both polluting and non-polluting

factors.

The study reveals compelling insights into the factors aecting eciency and productivity within Ecuador’s

private oil companies. Notably, we observe a signicant inuence of company size and technological

change, particularly among rms employing more polluting inputs in their production processes. Over

the study period, on average, companies display limited positive changes in eciency and productivity,

underscoring the need for targeted public policies aimed at reducing energy consumption in these rms.

Furthermore, consideration of electricity subsidies may incentivize more ecient and environmentally

conscious consumption practices.

This research highlights the pivotal role of energy eciency and environmental productivity in the oil

sector’s sustainability eorts. The ndings emphasize the necessity for proactive public policies to

curb energy consumption within private oil companies in Ecuador, aligning economic growth with

environmental responsibility. These insights are invaluable for policymakers and industry stakeholders

striving to strike a balance between protability and ecological stewardship within the Latin American oil

industry, with Ecuador serving as a pertinent case study.

Keywords: energy, eciency, productivity, environmental productivity, oil, companies.

Abstract

According to the International Energy Agency, the

oil industry contributes to approximately a third

of the world’s total carbon emissions (IEA 2021).

Thus, oil companies must become more ecient

and balance pollution mitigation and economic

performance. Some studies show the importance

of energy eciency in improving the economic

performance of oil companies by reducing costs

(Midor, et al. 2021, Yáñez, et al. 2018, Longwell

2002). However, when assessing the energy

eciency of oil companies, most studies have

frequently ignored environmental aspects (Hou, et

al. 2019, Jung, Kim and Rhee 2001). Therefore,

fewer studies are focusing on the environmental

performance of oil companies. According to the

literature in production economics, environmental

productivity refers to the ecient utilization of

pollution abatement and how this might inuence

the costs of alternative production and pollution

abatement technologies (Kaneko and Managi

2004). Studies in this eld are scarce, and most

have been developed in developed countries and

Asia.; (see, e.g., Tavana et al. (2019), Wegener and

Amin (2019), Sueyoshi and Wang (2014, 2018),

Da Silveira et al. (2017), Azedeh et al. (2015),

Song et al. (2015), Sueyoshi and Goto (2015),

among others). To the author’s knowledge, no

studies have been developed in which energy

eciency and environmental productivity change

in the oil sector is evaluated in Latin America, nor

has a specic case study been done on the oil

sector in one country in the region. Therefore,

the research problem focuses on “How is energy

eciency related to environmental productivity in

the Latin American oil sector, and how do these

variables impact the economic performance of oil

companies in a specic country within the region?”

This study aims to address the gap in the

academic literature by examining the relationship

between energy eciency and environmental

productivity within the Latin American oil industry

and assessing their impact on the protability of

oil companies in a specic context. Furthermore,

it seeks to contribute to the knowledge base on

industrial-level energy eciency analysis within

a developing nation. Specically, the research

1 INTRODUCTION

objective is to investigate the operational

dynamics of drivers and barriers inuencing energy

eciency in Ecuador’s industrial sector. Through

empirical investigation, this study will shed light

on the resource utilization practices of private oil

companies in this South American country, with a

particular focus on energy resources. Ultimately,

the primary goal is to provide valuable insights that

can help oil companies optimize resource usage,

enabling them to maximize prots while reducing

their environmental emissions.

For this study, it was considered a sample of

18 Ecuadorian private oil companies associated

with crude oil extraction and rening activities

in Ecuador was considered. Ecuador is the

fth oil producer in South America. In 2019 oil

extraction was 193.8 million barrels, of which

40.96 million barrels (21%) were extracted by

private companies. Among all industry sectors,

the petroleum industry is of particular interest

to Ecuador because of its economic and

environmental signicance. Public and private

companies own the oil industry in Ecuador. The

public sector plays a more signicant role due to

more production and higher investment (World

Bank 2018). Although, between 2000 and 2006,

the sector was led by private investment. A shift in

contract agreements in 2011 resulted in a decrease

in the investment made by private operators.

Oil is also essential for the Ecuadorian energy

sector; in 2018, Oil represented 86.9 percent of

the national energy supply. According to the Third

National Communication on Climate Change and

First Biennial Update Report (UNFCCC 2017), in

Ecuador, the energy sector produced 37 594 Gg

of carbon dioxide equivalent (CO2e), representing

47 percent of total GHG emissions in 2012. The

energy industry is a signicant contributor to GHG

emissions in the country, especially for the burning

of fossil fuels. In 2012 this activity accounted for

36 822.54 Gg (CO2e), representing 97.95 percent

of energy sector emissions.

Based on production value added during 2011-

2020, the following sectors had the most

signicant share in GDP: Manufacture (14.10%),

2 LITERATURE REVIEW

In a context where natural resources are

increasingly constrained, it is important to consider

that a company’s environmental productivity (EP)

is an essential piece of information that companies

needs to contemplate when they want to improve

their performance. It is helpful to review what is

meant by the term “productivity.” Productivity

expresses a relationship between the quantity of

goods and services produced by a business, or an

economy and the quantity of labor, capital, energy,

and other resources needed to produce those

goods and services (Finman & Laitner, 2001).

Meanwhile, EP involves the analysis of a company’s

relative eciency in its use of and impact on natural

resources (Wang & Shen, 2016). According to the

National trade (10.50%), Agriculture and shing

(9.18%), and Oil and quarrying (8.53%). Also, in

the period analyzed, oil exports accounted for

54.83% of total exports, and oil revenues for 30%

of overall scal income (Central Bank of Ecuador

2021).

To assess environmental eciency and

environmental productivity in Ecuador’s oil

companies, a non-parametric production model

(Tulkens 1993) is applied as a practical approach

to evaluating the pollution-adjusted productivity

change of Ecuadorian petroleum companies.

This method is widely applied in the literature

for production analysis (Sueyoshi, Yuan and

Goto 2017, Zhou, Ang and Poh 2008). Unlike

parametric models, this type does not require

explicitly specifying a mathematical form for

the production function. Moreover, it allows for

assessing the environmental eciency of multi-

inputs and multi outputs production units by

relaxing the convexity property of the pollution-

generating technologies. To the best of the

author‘s knowledge, no research has been

performed in the oil industry eld that analyses

environmental productivity change considering a

pollution-generating production model. Knowing

the prominent drivers of energy eciency and

environmental productivity change is a signicant

concern in the applied economics literature (Miao,

et al. 2019, Shen, Boussemart and Leleu 2017,

Valadkhani, Roshdi and Smyth 2016) This chapter

displays the main components of the pollution-

adjusted productivity variation considering

Ecuadorian oil companies. Identifying the primary

sources of pollution-adjusted productivity change

allows for displaying internal (technological

processes, management skills, Etc.) or external

(environmental policies, economic context, etc.)

constraints that inuence productivity variation.

The results suggest eciency and productivity

losses relate to energy consumption levels and

lack of technical change during the period.

The remainder of this research is structured

as follows. Section 2 displays the studies that

approach the driver of energy eciency and

the non-parametric models to estimate energy

eciency. The parametric and non-parametric

approach is presented in Section 3. The empirical

illustration is provided in Section 4. Finally, Section

5 focuses on the discussion and conclusions of

this research.

2.1.1 Environmental productivity

literature in production economics, environmental

productivity refers to ecient utilization of pollution

abatement and how this might inuence the costs

of alternative production and pollution abatement

technologies (Kaneko & Managi, 2004). Studies

related to environmental productivity are scarce,

and most have focused on developed countries

(Beltrán-Esteve, Giménez, & Picazo-Tadeo, 2019)

and Asia (Kaneko & Managi, 2004).Most studies

reviewed focus on implementing environmental

regulation to improve environmental productivity

in companies and countries (Wang & Shen, 2016;

Dewar, 1984). Also, some of these issues are

widely covered over industrial energy eciency.

studies in this eld have found that improving

energy eciency and incorporating energy

eciency technologies have signicant benets

on environmental productivity and allows to meet

sustainable development goals (Cagno, Worrell,

Trianni, & Pugliese, 2013).

Some studies review the relationship between

energy eciency improvement measures and

productivity in the industry. Finman & Laitner

(2001) reviewed more than 77 industrial case

studies. the authors suggest that energy eciency

investments yield signicant non-energy benets,

which are often not calculated. The description of

energy-ecient technologies as opportunities for

larger productivity improvements has signicant

implications for re-thinking how we quantify the

savings associated with capital investment and

the leverage points for promoting energy eciency

but may even challenge methods to use for

conventional economic assessments. Blumstein

et al. (1980) identies six kinds of barriers that

rms face to achieving industrial energy eciency:

1) misplaced incentives, meaning the economic

gains of obtaining energy eciency are not

always perceived by the decision makers. 2) lack

of information. 3) regulation. referring to existing

legal framework that conicts with cost-eective

measures. 4) market structure. as for example,

the energy eciency solution is not oered on the

market. 5) nancing, such as technologies that

requires high initial investment. 6) rm’s customs,

as company practices that generate low energy

eciency performance. However, when assessing

Knowing the primary sources of eciency and

productivity variation is of particular interest in the

economic literature. Non-parametric programming

modelings for production analysis are broadly applied

to assess these issues. Some studies employed

a DEA methodology using linear programming

techniques (Boussoane, Dyson, & Thanassoulis,

1991) to deal with undesirable outputs, such as

GHG emissions, which ultimately aect companies’

eciencies. Many approaches have been put forward

2.1.2 Energy Eciency and environmental productivity

2.1.3 Energy eciency and environmental productivity estimation methods

energy eciency and industry productivity, most

studies have frequently ignored environmental

aspects to improve productivity (Jung, Kim, &

Rhee, 2001). In addition, few studies focus on

the environmental performance of oil companies

(Hou, et al., 2019).

In the case of developing countries, the adoption

of energy eciency technologies and better

practices with clear sustainable goals by rms

are rarely explored in the literature. One of the

reasons may be the lack of management support,

prioritizing growth over environmental protection

(Grover & Karplus, 2020). The ndings of Karplus,

Shen, and Zhang (2020) suggest that companies

in China do not usually consider energy eciency

interventions with return periods longer than one

year. Energy eciency eorts are essential in

improving processes, minimizing the Impacts of

oil quality depletion, and achieving sustainable

development (Keskin, Dincer, & Dincer, 2020).

Aordable clean energy and climate action are

among the seventeen sustainable development

goals. Energy security and environmental

protection have become one of the most important

issues on today’s international agenda.

to account for this issue, such as parametric output

and input distance functions (Färe, Grosskopf, Knox,

& Yaisawarng, 1993; Coggins & Swinton, 1996;

Hailu & Veeman, 2001; Ho, Dey, & Higson, 2006)

and DEA methods (Skevas, Lansink, & Stefanou,

2012; 2014; Serra, Chambers, & Lansink, 2014;

Kabata, 2011; Yang, Wei, & Chengzhi, 2009; Ramli,

Munisamy, & Arabi, 2013).

Song, Zhang, and Wang (2015) applied the

2.2 Methodological Framework

2.2.1. Non-parametric model: DEA model and environmental productivity adjusted Malmquist

Index.

Network DEA model to divide eciency scores

into two subcategories, thus feeding back

more accurate results. In China, production and

environmental eciency changes were evaluated

in twenty local oil companies. Sueyoshi and

Goto (2015) incorporated Malmquist’s index in

the environmental assessment of oil companies’

studies. Azedeh, Mokhtari, Sharabi, and Zarrin

(2015) demonstrated the usability of DEA in studies

related to health, safety, and the environment in

an oil renery, improving ergonomic features in

To analyze the issue of energy eciency and

environmental productivity in private oil companies

in Ecuador, this research employs a DEA model.

DEA is an eciency evaluation method based on

the concept of relative eciency. There are dierent

types of DEA model such as SMB—DEA model,

that is non-radial and non-input or non-output

oriented, directly utilizes inputs and outputs to

determine the eciency measurement of DMUs.

In line with this study’s purpose, the SMB—DEA

model with undesirable output is applied to

estimate the energy eciency and environmental

productivity of 18 private oil companies in

Ecuador. This study only incorporates variables

whose values can be changed in a reasonable

period by decision-making units (Çelen, 2013),

This section displays the eciency evaluation

and productivity indices. The DEA method takes

an economic system or a production process

as an activity, where an entity (a unit) produces

a certain number of “productions” by investing

a certain number of elements within a limited

range (Li, Li, & Wu, 2013). These entities (units)

are called decision-making units (DMUs). Many

DMUs constitute to be respective evaluation

groups. The ecient production frontier is built on

evaluating, with each input or output indicator’s

weight as the variable under the analysis of input

and output ratios. In the end, an ecient DMU or

an inecient DMU can be determined according

to the distance between this DMU and the ecient

the business. Tavana et al. (2019) dened a fun

multi-objective multi-period network DEA model

customized to evaluate the dynamic performance

of oil reneries in the presence of undesirable

outputs. Considering the above, this empirical

study proposed a non-convex DEA modelling

and a parametric model to analyze oil industry

energy eciency and productivity with undesirable

outputs in private companies in Ecuador.

and that allows for maximizing the benets of oil

extraction and minimizing undesirable outputs.

To study and compare the dynamic eciency

of energy productivity among oil companies the

Malmquist Productivity Index (MPI) is adopted.

The MPI approach assesses the multi-faceted and

multi-output environmental impact of time frame

changes. This approach is used to account for

the change in industry policy eciency, with the

advantage of estimating the functional association

betweeninputs and outputs. The Malmquist and

DEA approach are among the most used tools to

estimate energy eciency in industry (Zhou, Ang,

& Poh, 2008; Zheng, 2021). These methods are

presented in more detail in the following sections.

production frontier (Debreu, 1951; Farrell, 1957;

Shephard, 1953). These distance functions fully

multiple inputs-outputs production processes.

The following denition presents the multiplicative

distance function (Abad, 2018).

Deﬁ nition 1.

The multiplicative pollution adjusted function

is employed to compute the Malmquist index.

According to Nishimizu and Page (1982), this

index can be discomposed into technical change

(TEC) and technical e ciency change (EC) when

examining productivity change. TC was de ned

as change in the best practice production frontier,

If the e ciency changes in is greater

than 1 then, e ciency progress arises over the

periods (t) and (t + 1). Moreover, technological

improvement occurs between the periods (t) and

(t + 1) when

Where:

are outputs and inputs

are the distance functions

between

vectors in

while EC was de ned to include all other productivity

change, including ‘learning by doing, di usion of

new technological knowledge, improved managerial

practice, scale e ciency and so on’.

The next equations display the productivity index for

the model:

If the e ciency changes in is greater

than 1 then, e ciency progress arises over the

2.2.2. Parametric model: Panel regression

We investigated the relationship between

productivity index and economic variables using a

Tobit panel regression model to specify individual

DMU e ects and cross-section data commonalities

(Liu & Liu, 2016). The standard linear model is

not appropriate for such analysis, because the

predicted values of e ciency scores may lie

outside the unit interval. As the accumulation of

scores at unity is a natural consequence of the

DEA approach, the Tobit model was employed

(Riaño & Larres, 2021).

The relationship between energy practices and oil

companies and the e ciency score is described

using the model below:

Where MI is the dependent variable, representing

the scores obtained from the eciency evaluation.

Emissions represents CO2 emissions per capita,

introduced in logarithms and Capital in level,

measured by the capital to labor ratio. Employment

and is the labor, measured in person, and Energy is

energy consumption measure in kwts/hour.

A sample of 18 private oil companies in Ecuador

is considered over the period 2011–2020. The

data set used in this research is built with the

population of registered oil Ecuadorian formal rms,

constructed from the balance sheets and nancial

statements registered on the ocial website of the

Superintendencia de Compañias, Valores y Seguros

(SCVS). This information is reported annually directly

by rms to the SCVS.

The inputs and outputs selected are used in

other DEA studies before for eciency analysis of

energy related industries to assess and monitor

technical eciency performance across a sample of

companies, these inputs and outputs are directed

related to the production process and have a greater

relevance on the enterprises management level

(Perreto et al. (2022).. Three inputs are selected: (i)

number of formal employees of each company and

(ii) net tangible assets (capital stock). Information

about the number of legally registered employees (i)

is declared by each company. The capital stock (ii)

is set as the sum of the real dollar value of buildings,

machinery and vehicles by assuming a depreciation

of 5, 10, and 20 percent. Precisely, the methodology

of Camino-Mogro and Bermudez-Barrezueta (2021)

is employed. Hence, the capital stock is valued

2.3 Data in brief

considering the gross investment in equipment in

year (t), net xed assets in real value (physical capital

in year (t – 1)), a depreciation rate and the price index

for equipment at the industry level obtained from the

Ecuadorian National Institute of Statistics. And, the

energy consumption of rms, measure in kilowatts/

hour, that considers the energy consumption of fossil

fuels registered by rms in the ocial statements

provided by SCVS. These in-puts permit to produce

dierent outputs. Thus, we consider one desirable

output, (iii) number of oil barrels and one undesirable

output represented by (iv) CO2 emissions.

The number of extracted barrels of oil (iii) is dened

based on the variable “sales” (American dollars)

reported in the balance sheets and nancial

statements registered on the ocial website of the

SCVS. Obviously, we divide it by the price (American

dollars/barrel) to obtain the variable “number of

extracted barrels of oil”. The reference price (WTI)

is considered allowing comparisons with another

international research in the same eld. The CO2

emissions (tons of CO2 equivalents) (iv) is measured

by using the methodology of the 2006 IPCC

Guidelines for National Greenhouse Gas Inventories.

Table 1. Characteristics of inputs and outputs

Variables Min Max Median S.D. Mean

Labor

0 6.55 2.30 2.06 2.73

7.47 18.97 13.26 2.12 13.48

Capital stock

(constant)

8.14 19.85 15.64 2.89 14.89

Energy

Consumption

5.95 16.44 12.89 2.30 12.27Oil production

1.31 22.41 8.79 4.93 9.75CO emissions

Source: Author, Notes: All variables in logarithms

Table 1. Characteristics of inputs and outputs

Table 1 presents the descriptive statistics of

the variables used in this study. The statistical

description of the data set displays variation

in the database. The standard deviation (S.D.)

This table represents the correlation matrix for

the input and output variables in the sample. The

variables selected as inputs are highly correlated

with the outputs conferring validity to our empirical

To study energy eciency for oil companies

in Ecuador, this research used the SMB-DEA

model to consider for undesirable output. This

analysis presents two scenarios. In scenario 1,

energy inputs and outputs are involved in the

production of good and bad outputs. In contrast,

scenario 2 only considers energy input to produce

the desirable output. The results of these two

scenarios—Technical-factor energy eciency

In a DEA model the companies whose eciency

is 1 or greater than 1 make up the production

frontier compared to those whose eciency is

less than 1, which are DEA inecient. Table 3

reveals that in Scenario 1 (the production function

with undesirable and o desirable outputs), only 4

companies showed ineciency scores. On the

other hand, in Scenario 2, 6 rms registered an

2.3 Correlation matrix

3.1. Analysis of technical-factor energy eciency (TFEE) and Particular-factor energy

eciency (PFEE)

values indicate unbalanced growth of private oil

companies in Ecuador over the period 2012-2020.

strategy. The high correlation found also conrms

the association between the selected inputs and

outputs as statistically signicant at 90%.

3 RESULTS

(TFEE) and Particular-factor energy eciency

(PFEE) allows a deeper exploration of energy

eciency in extraction incentive industries. Then

the Malmquist index productivity is calculated

to understand the change in energy productivity

across the time period. Additionally, a Tobit panel

regression is conducted to analyze the potential

drivers of energy eciency for these Ecuadorian

oil rms.

energy productivity scores less than one. Thus,

these results are consistent with the ndings of

Wang et al.(2019) and Tachega et al.(2020), who

suggest that a production function that integrates

energy and traditional economic inputs can

increase oil production and reduce CO2 emissions

with overall good eciency score levels.

Variables

Energy

consumption

Employment CapitalO il production

Energy Consumption1

0.0285

0.4267***

0.7483***

0.2839***

0.4442***

0.7483***

0.1361*

0.5080*** 1

0.2045*** 0.5132*** 1

1Employment

Capital

Oil production

CO emissions

Source: Author , Notes: *p<.1, **p<.05, ***p<.01

Table 2. Correlation Matrix

3.2. Malmquist Index pollution-adjusted productivity

The results outlined in the table 4 reveal the

PM productivity indices scores and their

decompositions over the period 2011-2020. The

rst column displays the Malmquist index scores

(MC), and the other two columns show the main

drivers of the environmental productivity change,

namely the technological change (TC) and the

eciency variation components (EC), respectively.

Source: Author

Table 3. Energy eciency scores for TFEE and PFEE

Table 4. Malmquist Index scores for 2012-2020

AMODAIMI-OIL COMPANY. S.L.

ANDES PETROLEUM ECUADOR LTD.

CARLOS PUIG & ASOCIADOS S.A. CIA.

DE EXPLORACION DE MINERALES Y SERVICIOS MINEROS

COMPAÑIA SUDAMERICANA DE FOSFOROS DEL ECUADOR

FOSFOROCOMP S.A.

ENAP SIPETROL S.A.

EQUIPENINSULA S.A

EQUIPO PETROLERO S.A. EQUIPETROL

ERINCORP S.A.

HILONG OIL SERVICE & ENGINEERING ECUADOR CIA. LTDA.

LOGISPETROL SERVICIOS PETROLEROS CIA. LTDA.

OVERSEAS PETROLEUM AND INVESTMENT CORPORATION

PDVSA ECUADOR S.A.

PETROLEOS SUD AMERICANOS DEL ECUADOR PETROLAMEREC S.A.

PETROORIENTAL S.A.

PETRORIVA S.A.

REPSOL ECUADOR S.A.

SAXON ENERGY SERVICES DEL ECUADOR S.A.

TECPECUADOR S.A.

1.091

0.383

1.813

1.017

0.53

2.515

0.611

1.22

0.584

1.832

1.21

0.612

1.471

0.965

1.045

0.469

1.331

1.16

1.163

1.122

0.648

1.426

1.099

1.087

0.84

1.096

0.919

0.825

1.34

0.72

1.0

0.97

1.129

1.49

0.72

1.034

SCENARIO2SCENARIO 1

AVERAGE 1.103 1.037

3.2.1. Analysis of overall eciency (MI)

3.2.2. Analysis of technical and eciency variation changes

3.3. Tobit Panel Regression results

Table 4 reports the average annual PM productivity

indices for the 18 oil companies in Ecuador over

the analyzed period. In the DEA model, the

companies whose eciency is 1 or greater than

1 make up the production frontier, compared to

those whose eciency is less than 1 which are

DEA inecient. Therefore, the results in Tables 4

for the overall energy eciency (MI) score showed

that more than half of the companies are inecient

during the time frame. The group of companies

have an average of energy eciency score of

1.80. From this group, only 3 companies have

a higherMalmquist Index Score than the average.

In order words, only three rms perform better

than the average. The slowdown in productivity

The mean technical eciency change (TC) for the

18 companies selected in the period analyzed was

- 0,091%, meanwhile there was not a signicant

scale change (EC) over time. Globally, the results

suggest that the energy eciency performance of

the Ecuadorian oil industry is dependent on the

technical change in production, but it is important

to note:

Having obtained the PMI analysis, we want to

nd the primary economic indicators that aect

eciency scores. The Hausman test

is employed

to choose between the xed-eect and random-

eect model—suitable for the panel regression

1. In relation to the overall energy eciency

scores for 2011-2012, 2012-2013 and

2014-2015, most companies presented

a drop in the technical and eciency

component scores during the period

analyzed. This means that the energy

ineciency of these rms was driven by

scores could be linked to rms with higher levels

oil and gas production and CO2 emissions during

the analyzed period, as most rms with low

consumption of fossil fuels have a better ratio

between output and pollution, and consequently,

are more sustainable. On the other hand, the energy

eciency scores for most companies exhibit an

important decrease between 2012-2019 as seen

in gure 2.1., this period coincides with important

reforms in Ecuador referring to private contribution

in the oil sector, resulting in lower investment in

capital projects and less resources designated

for innovation in these companies (World Bank,

2018).

less technological advances without any

commensurate eciency improvements in

the internal management of the rms.

2. For 2018-2019 the PMI index show

marginally reduce and a then a positive

boost in 2019-2020, these results suggest

that although in 2020 the industry suered

an important reduction in oil production due

to the Covid-19 outbreak, the overall energy

eciency and productivity levels were

positive aected, and that could be related

to the decrease in CO2 emissions during

the period even if there weren’t signicant

technical and energy eciency change.

analysis. The results indicate the random eect

model is more suitable for the panel regression

evaluation.

3.- The test proposed by Hausman Invalid source specied. is a chi-square test that determines whether dierences are systematic and

signicant between two estimates. It is mainly used to determine whether an estimator is consistent or whether a variable is relevant or not.

Thus, in the next step, we employ the random eect

model to measure the impact of the indicators on

PMI (Table 5.). Per the analysis, MI has a weak

negative correlation with energy consumption at a

10% signicance level. And a negative relationship

with employment at a 1% signicance level. These

results suggest that for Ecuador, the energy and

industrial eciency of oil companies depends on

their labor strategy and the consumption of fossil

fuels in their extractive activities.

Source: Author

Table 5. Panel regression results

Variables

Energy consumption

Employment

Capital

Oil production

CO2 emissions

Observations

Number of n

-0.0193*

(0.0103)

-0.138***

(0.0210)

-0.0139

(0.0175)

0.0508**

(0.0229)

-0.00334

(0.00307)

180

The objective of this study was to analyze the

main drivers of eciency and productivity in

private oil companies in Ecuador, with a particular

focus on energy eciency and its relationship with

environmental factors.

We conducted this analysis using a dataset

comprising 18 Ecuadorian oil companies over the

period 2011-2020. To evaluate energy eciency

and productivity, we employed a non-parametric

model and the Malmquist index, allowing us to

assess both pollution-adjusted productivity and

the factors contributing to eciency.

Our analysis revealed that more than half of the

companies in our study were characterized as

inecient based on the DEA model. The average

energy eciency score of 1.80 underlines the

industry-wide challenges in achieving optimal

energy eciency. This trend was particularly

notable among companies with higher levels

of oil and gas production and associated CO2

emissions during the analyzed period. In contrast,

companies with lower fossil fuel consumption

demonstrated a more favorable output-to-pollution

ratio, highlighting their greater sustainability in

terms of energy eciency.

The declining energy eciency scores observed for

most companies from 2012 to 2019 coincide with

signicant reforms in Ecuador’s oil sector. These

reforms led to reduced investments in capital

projects and innovation within these rms, as

reported by the World Bank (2018). This suggests

that policy shifts can have a substantial impact on

energy eciency levels within the industry.

The Malmquist index scores (MC), when

decomposed into technological change (TC) and

eciency variation components (EC), provide

valuable insights into the industry’s performance.

The analysis indicates minimal overall changes

in scale over time, emphasizing the industry’s

dependence on technical advancements in

4 CONCLUSIONS

production to drive energy eciency improvements.

Two critical observations emerge from our ndings.

During the periods 2011-2012, 2012-2013,

and 2014-2015, both technical and eciency

component scores declined. This indicates that

energy ineciency during these periods was

primarily driven by a lack of technological progress

without corresponding eciency improvements in

internal management.

In 2018-2019, there was a marginal reduction

in the PMI index, followed by a positive boost

in 2019-2020. This suggests that, despite a

signicant reduction in oil production due to the

Covid-19 outbreak in 2020, energy eciency and

productivity levels improved. This improvement

may be attributed to decreased CO2 emissions,

even in the absence of signicant technical and

energy eciency changes.

Correlation Analysis: To further understand

the factors inuencing eciency scores, we

conducted a panel regression analysis using the

random eect model. The results indicated a weak

negative correlation between energy consumption

and MI at a 10% signicance level. Additionally,

employment displayed a negative relationship

with MI at a 1% signicance level. This implies

that energy and industrial eciency in Ecuador’s

oil companies are closely linked to their labor

strategies and fossil fuel consumption in extractive

activities.

In conclusion, our study has highlighted the

formidable energy eciency challenges faced by

Ecuador’s oil industry, with signicant implications

for environmental sustainability and protability.

Policy reforms, technological progress, and internal

management practices all play pivotal roles in

shaping energy eciency outcomes. The ndings

underscore the industry’s need for comprehensive

management strategies that address both human

resources and resource utilization. As the sector

navigates evolving challenges, the imperative

to prioritize sustainability and eciency remains

paramount for achieving a harmonious balance

between economic growth and environmental

responsibility.

5 BIBLIOGRAPHY

Abad, A. (2018). Les enseignements de la micro-économie de la production face aux enjeux environnementaux :

étude des productions jointes théorie et applications. Université de Perpignan. Retrieved from https://tel.archives-

ouvertes.fr/tel-01963415.

Azadeh, A., Raoo, Z., & Zarrin, M. (2015). A multi-objective fuzzy linear pro- gramming model for optimization

of natural gas supply chain through a greenhouse gas reduction approach. Journal of Natural Gas Science and

Engineering, 26, 702-710. doi:https://doi.org/10.1016/j.jngse.2015.05.039.

Banker, R., & Maindiratta, A. (1986). Piecewise loglinear estimation of ecient production surfaces. Management

Science, 32(1), 126–135. doi:https://doi.org/10.1287/mnsc.32.1.126

Beltrán-Esteve, M., Giménez, V., & Picazo-Tadeo, A. (2019). Environ- mental productivity in the European Union:

A global Luenberger-metafrontier approach. Science of The Total Environment, 692, 136-146. doi:https://doi.

org/10.1016/ j.scitotenv.2019.07.182.

Blumstein, C., Krieg, B., Schipper, L., & York, C. (1980). Overcoming social and institutional barriers to energy

conservation. Energy, 5(4), 355-371. doi:https://doi.org/10.1016/0360-5442(80)90036-5.

Boussoane, A., Dyson, R., & Thanassoulis, E. (1991). Applied data envelopment analysis. European Journal of

Operational Research, 52(1), 1-15. doi:https://doi.org/10.1016/0377-2217(91)90331-O.

Cagno, E., Worrell, E., Trianni, A., & Pugliese, G. (2013). A novel approach for barriers to industrial energy eciency.

Renewable and Sustainable Energy Reviews, 19, 290-308. doi:https://doi.org/10.1016/j.rser.2012.11.007.

Camino-Mogro, S., & Bermudez-Barrezueta, N. (2021). Productivity determinants in the construction sector in

emerging country: New evidence from Ecuadorian rms. Review of Development Economics, 25(4), 2391–2413.

doi:https://doi.org/ 10.1111/rode.12771.

Çelen, A. (2013). Eciency and productivity (TFP) of the Turkish electricity distribution companies: An application

of two-stage (DEA&Tobit) analysis. Energy Policy, 63, 300-310. doi:https://doi.org/10.1016/j.enpol.2013.09.034.

Central Bank of Ecuador. (2021). Reporte del Sector Petrolero IV Trimestre de 2021. Technical report. Retrieved

from https://contenido.bce.n.ec/documentos/Estadisticas/Hidrocarburos/ASP202104.pdf

Coggins, J., & Swinton, J. (1996). The price of pollution: A dual approach to valuing SO2 allowances. Journal of

Environmental Economics and Management, 30(1), 58–72. doi:https://doi.org/10.1006/jeem.1996.0005.

Da Silveira, P., Callegari, C., Ribas, A., Lucena, A., Portugal-Pereira, J., Koberle, A., . . . Schaeer, R. (2017).

The power of light: Socio-economic and environmental implications of a rural electrication program in Brazil.

Environmental Research Letters, 12(9). doi:https://doi.org/10.1088/1748-9326/aa7bdd

Debreu, G. (1951). The coecient of resource utilization. Econometrica, 19(3), 273–292. doi:https://doi.

org/10.2307/1906814.

Dewar, R. (1984). Environmental Productivity, Population Regulation, and Carrying Capacity. American

Anthropologist, 86(3), 601-614. doi:https://doi.org/ 10.1525/aa.1984.86.3.02a00040.

Färe, R., Grosskopf, S., Knox, C., & Yaisawarng, S. (1993). Derivation of shadow prices for undesirable outputs:

A distance function approach. The Review of Economics and Statistics, 75(2), 374-380. doi:https://doi.

org/10.2307/2109448

Farrell, M. (1957). The measurement of productive eciency. Journal of the Royal Statistical Society, 120(3),

253–281. doi:https://doi.org/10.2307/2343100.

Finman, H., & Laitner, J. (2001). Industry, energy eciency and productivity improvements. In Proceedings of the

2001, Summer Study on Energy Eciency in Industry. 561-570.

Grover, A., & Karplus, V. (2020). The energy-management nexus in rms : Which practices matter, how much and

for whom? doi:https://doi.org/10.1596/ 1813-9450-9397.

Hailu, A., & Veeman, T. (2001). Alternative methods for environmentally adjusted productivity analysis. Agricultural

Economics, 25(2-3), 211-218. doi:https://doi.org/10.1016/S0169-5150(01)00078-0.

Ho, W., Dey, P., & Higson, H. (2006). Multiple criteria decision-making techniques in higher education. International

Journal of Educational Management, 20(5), 319–337. doi:https://doi.org/10.1108/09513540610676403.

Hou, Y., Iqbal, W., Muhammad, G., Iqbal, N., Ahmad, Y., & Fatima, A. (2019). Measuring Energy Eciency and

Environmental Performance: A Case of South Asia. Processes, 7(6), 325. doi:https://doi.org/10.3390/ pr7060325.

IEA. (2020b). World energy outlook 2020.

IEA. (2021). Oil 2021 - Analysis and forecast to 2026 (Technical report). Paris: International Energy Agency.

Retrieved from https://iea.blob.core.windows.net/assets/1fa45234-bac5-4d89-a532-768960f99d07/Oil_2021-

PDF.pdf

Jung, E., Kim, J., & Rhee, S. (2001). The measurement of corporate environmental performance and its

application to the analysis of eciency in oil industry. Journal of Cleaner Production, 9(6), 551-563. doi:https://doi.

org/10.1016/S0959-6526(01)00011-7

Kabata, T. (2011). The US agriculture greenhouse emissions and environmental performance. Agricultural and

Applied Economics Association, 1–24. doi:https://doi.org/10.22004/AG.ECON.103427.

Kaneko, S., & Managi, S. (2004). Environmental Productivity in China. Economics Bulletin, 17(2), 1-10.

Karplus, V., Shen, X., & Zhang, D. (2020). Herding cats: Firm non-compliance in China’s industrial energy eciency

program. Energy JournalEnergy Journal, 41(4), 1-30. doi:https://doi.org/10.5547/01956574.41.4.vkar

Keskin, A., Dincer, B., & Dincer, C. (2020). Exploring the Impact of Sustainability on Corporate. Sustainability, 1-14.

Retrieved from le:///C:/Users/CAAP%20BACKUP%201/Downloads/sustainability-12-02346-v2.pdf

Li, L., Li, M., & Wu, C. (2013). Production eciency evaluation of energy companies based on the improved super-

eciency data envelopment analysis considering undesirable outputs. Mathematical and Computer Modelling,

58(5-6), 1057-1067. doi:https://doi.org/10.1016/j.mcm.2012.07.001

Liu, X., & Liu, J. (2016). Measurement of low carbon economy eciency with a three- stage data envelopment

analysis: A comparison of the largest twenty CO2 emitting countries. International Journal of Environmental

Research and Public Health, 13(11), 1116. doi:https://doi.org/10.3390/ijerph13111116

Longwell, H. (2002). The future of the oil and gas industry: past approaches, new challenges. World Energy, 5(3),

100-104.

Managi, S. (2011). Technology, natural resources and economic growth: Improving the environment for a greener

future. Edward Elgar. doi:https://doi.org/10.4337/9781849807418.

Miao, Z., Chen, X., Baležentis, T., & Sun, C. (2019). Atmospheric environmental productivity across the provinces

of China: Joint decomposition of range adjusted measure and Luenberger productivity indicator. Energy Policy,

132, 665-677. doi:https://doi.org/10.1016/j.enpol.2019.06.019

Midor, K., Ivanova, T., Molenda, M., Biały, W., & Zakharov, O. (2021). Aspects of Energy Saving of Oil-Producing

Enterprises. Energies, 15(1), 259. doi:https://doi.org/10.3390/en15010259.

Nishimizu, M., & Page, J. (1982). Total Factor Productivity Growth, Technological Progress and Technical Eciency

Change: Dimensions of Productivity Change in Yugoslavia. The Economic Journal, 92(368), 920–936. doi:https://

doi.org/10.2307/2232675

Perreto, C., Quiroga, F., & Alberto, C. (2022). Análisis de los factores que determinan la eciencia de los bancos

en Argentina. Un análisis DEA en dos etapas. Contaduría y administración, 67(1), 375-396. Retrieved from https://

dialnet.unirioja.es/servlet/articulo?codigo=8387472

Ramli, N., Munisamy, S., & Arabi, B. (2013). Scale directional distance function and its application to the

measurement of eco-eciency in the manufacturing sector. Annals of Operations Research, 211(1), 381–398.

doi:https://doi. org/10.1007/s10479-013-1441-1.

Riaño, C., & Larres, O. (2021). Data envelopment analysis and applications in sustainability. Ingeniare, 17(31), 11-

19. Retrieved from https://dialnet.unirioja.es/descarga/articulo/8686310.pdf

Serra, T., Chambers, R., & Lansink, A. (2014). Measuring technical and environmental eciency in a state-

contingent technology. European Journal of Operational Research, 236(2), 706–717. doi:https://doi.org/10.1016/J.

EJOR.2013.12.037

Shen, Z., Boussemart, J., & Leleu, H. (2017). Aggregate green pro- ductivity growth in OECD’s countries.

International Journal of Production Economics, 189, 30-39. doi:https://doi.org/10.1016/j.ijpe.2017.04.007.

Shephard, R. (1953). Cost and production functions. Princeton University Press.

Skevas, T., Lansink, A., & Stefanou, S. (2012). Measuring technical eciency in the presence of pesticide spillovers

and production uncertainty: The case of Dutch arable farms. European Journal of Operational Research, 223(2),

550–559. doi:https://doi.org/10.1016/j.ejor.2012.06.034.

Skevas, T., Lansink, A., & Stefanou, S. (2014). Pesticide use, environmental spillovers and eciency: A DEA risk-

adjusted eciency approach applied to Dutch arable farming. European Journal of Operational Research, 237(2),

658–664. doi:https://doi.org/10.1016/j.ejor.2014.01.046.

Song, M., Zhang, J., & Wang, S. (2015). Review of the network environmental eciencies of listed petroleum

enterprises in China. Renewable and Sustainable Energy Reviews, 43, 65-71. doi:https://doi.org/10.1016/j.

rser.2014.11.050.

Sueyoshi, T., & Goto, M. (2015). DEA environmental assessment in time horizon: Radial approach for Malmquist

index measurement on petroleum companies. Energy Economics, 51, 329-345. doi:https://doi.org/10.1016/j.

eneco.2015.07.010

Sueyoshi, T., & Wang, D. (2014). Sustainability development for supply chain management in U.S. petroleum

industry by DEA environmental assessment. Energy Economics, 46, 360-374. doi:https://doi.org/10.1016/j.

eneco.2014.09.022.

Sueyoshi, T., & Wang, D. (2018). DEA environmental assessment on US petroleum industry: Non-radial approach

with translation invariance in time horizon. Energy Economics, 72, 276-289. doi:https://doi.org/10.1016/j.

eneco.2018.02.003.

Sueyoshi, T., Yuan, Y., & Goto, M. (2017). A literature study for DEA applied to energy and environment. Energy

Economics, 62, 104-124. doi:https://doi.org/10.1016/j.eneco.2016.11.006

Tavana, A., Emami, A., Houshfar, E., Mahmoudzadeh, A., Ashjaee, M., Shoaee, S., . . . Marashi, F. (2019). Toward

renewable and sustainable energies perspective in Iran. Renewable Energy, 139, 1194-1216. doi:https://doi.

org/10.1016/J.RENENE.2019.03.022.

Tulkens, H. (1993). On FDH eciency analysis: Some methodological issues and applications to retail banking,

courts, and urban transit. Journal of Productivity Analysis, 4, 183-210. doi:https://doi.org/10.1007/BF01073473

UNFCCC. (2017). First biennial update report of Ecuador. United Nations Framework

Convention on Climate Change. Retrieved from https://unfccc.int/sites/default/les/resource/

Presentation%7B%5C_%7DEcuador%7B%5C_%7Dnal.pdf

Valadkhani, A., Roshdi, I., & Smyth, R. (2016). A multiplicative environmental DEA approach to measure

eciency changes in the world’s major polluters. Energy Economics, 54, 363-375. doi:https://doi.org/10.1016/j.

eneco.2015.12.018.

Wang, Y., & Shen, N. (2016). Environmental regulation and environmental productivity: The case of China.

Renewable and Sustainable Energy Reviews, 62, 758-766. doi:https://doi.org/10.1016/j.rser.2016.05.048.

Wegener, M., & Amin, G. (2019). Minimizing greenhouse gas emissions using inverse DEA with an application in

oil and gas. Expert Systems with Applications, 122, 369–375. doi:https://doi.org/10.1016/J.ESWA.2018.12.058.

World Bank. (2018). Ecuador - Systematic Country Diagnostic. Technical report. doi:https://doi.org/10.1596/30052

Yang, H., Wei, Z., & Chengzhi, L. (2009). Optimal design and techno-economic analysis of a hybrid solar–

wind power generation system. Applied Energy, 86(2), 163-169. doi:https://doi.org/https://doi.org/10.1016/j.

apenergy.2008.03.008.

Yáñez, E., Ramírez, A., Uribe, A., Castillo, E., & Faaij, A. (2018). Unravelling the potential of energy eciency

in the Colombian oil industry. Journal of Cleaner Production, 176, 604-628. doi:https://doi.org/10.1016/j.

jclepro.2017.12.085

Zheng, Z. (2021). Energy eciency evaluation model based on DEA-SBM-Malmquist index. Energy Reports, 7,

397-409. doi:https://doi.org/10.1016/j.egyr.2021. 10.020.

Zhou, P., Ang, B., & Poh, K. (2008). A survey of data envelopment analysis in energy and environmental studies.

European Journal of Operational Research, 189(1), 1-18.