Journal of Cleaner Production 265 (2020) 121636

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Do voluntary environmental management systems improve

environmental performance? Evidence from waste management by
Kenyan firms
Edward M. Mungai a, S. Wagura Ndiritu a, Tazeeb Rajwani b, *
a
Strathmore University Business School, Kenya
b
University of Surrey, Surrey Business School, Department of Strategy and International Business, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: We examine whether the adoption of global voluntary environmental management systems - United
Received 5 December 2019 Nations Global Compact and ISO 14001 - lead to more effective environmental performance. Previous
Received in revised form studies have presented inconclusive findings of voluntary environmental management systems on
6 April 2020
environmental performances. Possible reasons for conflicting results are the influence of observable and
Accepted 10 April 2020
Available online 19 April 2020
unobservable factors that affect environmental performance as well as the use of different measures of
environmental performance indicators. Using primary data from Kenyan firms in 2019, waste manage-
Handling editor: Yutao Wang ment defined by wastewater recycling, solid waste reusing and use of environmentally safe disposal
methods, we determine the effects of voluntary environmental management systems (VEMS) on firm-
Keywords: level environmental performance. We conclude that the adoption of VEMS are associated with signifi-
Voluntary environmental management cant improvement in environmental performance in developing economies. Our conclusions provide
system insights to corporate management and policy makers in developing countries on decisions regarding
United Nations Global Compact better environmental management. VEMS forms an environmental management tool suitable in con-
ISO 14001
fronting wastewater and physical refuse challenges as well as conducive waste disposal methods.
Environmental performance and developing
© 2020 Elsevier Ltd. All rights reserved.
economies

1. Introduction industry and third-party organizations have developed and pro-

The world’s worst environmental concerns are in developing
countries which have compelled governments to establish laws
that require firms to cut down on pollution. These are command
and control approaches enforced by regulatory bodies. However, in
developing countries, environmental agencies lack funding,
expertise, and personnel that create insufficient resources for
environmental protection. Moreover, enforcing environmental
regulations has suffered gaps and inconsistencies compelling the
governments to seek alternatives policies that are not dependent
on mandates, monitoring compliance, and sanctions on violators. In
developing economies, solid waste management policies are
designed to cope with environmental issues but fail to incorporate
economic and health perspectives. Arimula et al. (2016) posit that
firms take these regulations as costly and in turn view waste as a
problem and not a recyclable resource. In response, governments,
* Corresponding author. Tel.: þ4401234751122.
E-mail address: t.rajwani@surrey.ac.uk (T. Rajwani).
moted the adoption of voluntary environmental management
systems (VEMS). Collectively, VEMS lay out a structure for an or-
ganization to identify and guide in enhancing its environmental
performance. Extensive scholarly research has been done on
assessing the success of VEMS in promoting environmental per-
formance in developed countries with less regard to the undevel-
oped economies (Lee et al., 2015; Nemati et al., 2016). Therefore,
our study seeks to assess if adoption of ISO 14001 and for the first
time United Nations Global Compact (UNGC) in developing coun-
tries lead to better environmental performance.
Waste Management is a major public health and environmental
concern in many developing countries. The situation in Africa,
particularly in the large urban towns is severe. The public sector is
unable to deliver services effectively, regulation of the private
sectors is limited, and illegal dumping of domestic and industrial
waste is a common practice. Institutions charged with the re-
sponsibility of providing waste management services have found it
increasingly challenging to play their role (UNEP, 2010).
This study will restrict itself to Kenya, a country that intends to
increase the contribution of manufacturing sector to the Gross

https://doi.org/10.1016/j.jclepro.2020.121636
0959-6526/© 2020 Elsevier Ltd. All rights reserved.
2 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636
Domestic Product (GDP) from 8% to 15% in 2022. This will require a
significant amount of natural resources consumption as well as
resulting to significant amount of waste (Government of Kenya,
2018). In Nairobi, less than half of the daily waste generated is
collected and lacks safe disposal methods with the majority ending
into indiscriminate dumpsites. Kenya’s waste management situa-
tion, which is largely characterized by low waste collection,
pollution from uncontrolled dumping of waste, inefficient public
services, unregulated and uncoordinated private sector and lack of
key waste management infrastructure. The expansion of industrial
and commercial sectors has resulted in increased urban migration,
improved standard of living and technological advancement which
in turn led to increased waste generation.
In response, the country adopted the National Solid Waste
Management Strategy as a guide to sustainable solid waste man-
agement to ensure a healthy, safe and secure environment for all.
This has culminated into a legislation on sustainable waste man-
agement, in form of a bill that is aimed to help Kenya achieve a
green economy characterized by zero waste. Kenyan as a green
economy is envisioned in the country’s Vision 2030 as well as in
line with the Kenyas National Determined Contribution (NDC)
commitment to the Paris Climate Agreement. The bill provides in-
centives aimed at the reduction of waste especially through the use
of more efficient technologies. There are also incentives in the areas
of sustainable product design, resource efficiency, re-using and
recycling of materials (Government of Kenya, 2018). Due to the poor
state of waste management in Kenya, the National Environmental
Management Authority (NEMA) adopted the minimum re-
quirements necessary for continuous improvement. They are based
on waste collection, transportation, disposal and licensing.
The impacts of poor solid waste management and used water
within the urban settlements, particularly cities and big munici-
palities can be disastrous. As such, there is a need for proper and
efficient waste management. Kenya Vision 2030 recognizes the
need for efficient and sustainable waste management systems to be
established as the country develops into a newly industrialized
state by 2030. Yong et al. (2016) affirm that waste management is
an important way of achieving cleaner and sustainable production.
Industries have opted for voluntary programs that provide extra
incentives such as subsidies and positive publicity with a win-win
outcome. In developed countries, voluntary programs are popular
and have been studied extensively. Less well known is the growing
environmental activities in Africa where firms are quickly
embracing voluntary approaches and rapidly adopting the initia-
tives. Voluntary environmental initiatives encourage beyond
compliance. Dingwerth (2008) posit these clean production ini-
tiatives hold considerable promise for poor countries. However,
Rusell and Vaughan (2003) identify weak institutions, incomplete
legal foundations and limited political will in developing countries
as constraints to voluntary program effectiveness in developing
countries. More and more companies are looking into enhancing
their environmental performance through the use of Environ-
mental Management System (EMS). This gives an opportunity to
examine how companies are engaging with EMS especially the
voluntary programs.
According to Khanna (2002), the threat of mandatory regulation
is a major driver for adopting voluntary initiatives. Consequently,
voluntary instruments perform dismally in countries where regu-
latory pressure is weak. The non-regulatory factors motivating
firms to enroll in voluntary initiatives such as consumers and
community groups are feeble in developing countries. A scenario
further complicated by weak environmental management bodies
such as advocacy groups and the private sector in developing
countries (Russell and Vaughan, 2003). Blackman (2010) charac-
terized developing economies as populous with small scale and
informal firms that are less prone to both regulatory and non-
regulatory pressure. While Seiffert (2008) suggests that small and
medium enterprises account for a large share of environmental
burden and especially regarding waste management and hence the
motivation for this study. Generally, the purpose of clean produc-
tion initiatives is to help institutions implement or improve their
EMS thereby improving their environmental performance
(Nawrocka and Parker, 2009). Despite that there are differing ex-
pectations of an EMS depending on its perception and role, a
possible measurable positive outcome is proper waste manage-
ment that would entail environmental performance.
Accreditation to the voluntary schemes translates into an inte-
grated multidisciplinary approach that not only includes clean
production technologies, but also green policies and programs
aimed at healthy ecosystems (Kleme
s et al., 2012). ISO 14001 offers
guidelines on the handling and treatment of waste both water and
solid. On the other hand, the UNGC addresses social and ecological
responsibilities within the organizational context. In prior studies,
ISO 14001 has been taken as a reference to discuss the legitimacy
and effectiveness of EMS (Rasche et al., 2013) leaving out UNGC.
Consequently, ISO 14001 and UNGC adoption have become popular
in addressing negative business impacts by advocating for standard
corporate social responsibility (Abbott and Snidal, 2010). A distinct
feature with voluntary programs is that they emphasize on pro-
cedures rather than clearly defined performance targets allowing
for flexibility but also provoking criticism.
This study makes several contributions to environmental man-
agement literature using evidence from the adoption of voluntary
environmental programs in Africa. Firstly, the conclusion to support
or dispute arguments based on the efficacy of EMS in developed
countries is indefinite (Kube et al., 2019). This is partly due to un-
explored spillover effects (Zhou et al., 2020), as well as the fact that
extant research has focused on exploring single VEMS - ISO 14001
(Lee et al., 2015; Nemati et al., 2016) and neglecting UNGC, an
equally popular VEMS. Majority of the past scholarship is on North
America and Europe (Liu et al., 2018), with much less on Asia or
Africa. The findings from industrialized countries may not apply in
developing economies due to the difference in some institutional
and socioeconomic contextual factors. We use firm level data on
Kenyan industrial and commercial firms, that constitute major in-
dustrial polluters, to examine the effectiveness of ISO 14001 and
UNGC.
Secondly, this novel study adds to the assessment of the extent
of application of clean production initiatives in emerging econo-
mies on holistic and preventative approaches towards their major
climate concern of waste management. This is achieved by
exploring the business environment both internal and external, to
clearly evaluate the effect of global EMS unlike a recent study by
Tatoglu et al. (2019) that extended earlier research explorations on
firm motivations to ratify with VEMS in Asia. Previous studies have
raised several pertinent questions on the purpose of self-regulatory
global governance initiatives, their goals and have tried to assess
their effectiveness and legitimacy from developed country per-
spectives (Rasche et al., 2013). To weigh in the debate, we assess the
effectiveness of ISO 14001 and UNGC, in the view of three major
environmental concerns in developing countries; solid waste
management, wastewater management and waste disposal
methods in Kenya. Decades of rapid urbanization, population
growth, and industrialization has seen developing countries con-
fronting severe air, water, and solid waste pollution concerns.
Kenya has experienced an increased waste generation, poor waste
collection services, open dumping and low levels of compliance
with environmental legislation. Major sources of solid waste in
Kenya are households and industries where Kenya generates 4
million tons of solid waste annually a figure set to double by 2030
 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636
(Government of Kenya, 2018). Unfortunately, the increase is not
accompanied by equivalent handling capacity.
The remainder of this paper proceeds as follows: The second
section provides a literature review. The third section describes the
research methodology. Section 4 discusses the main research
findings. Finally, in the last section conclusions and implications of
the research are provided.
2. Related literature and research hypotheses
Studies on UNGC focuses on three key areas: (1) the firms’
motivations to adopt the UNGC; (2) the composition of the UNGC
participants; and (3) the impact of committed firms’ operations and
business strategies. Regarding the latter, several works have
focused on analyzing the financial and non-financial benefits of the
UNGC adoption, such as improvements in corporate image, the
option of making the organization stand out from the competition,
learning better practices, obtaining competitive advantages, and
facilitating strategic changes when implementing the principles
(Arevalo et al., 2012). However, these studies paint a mixed picture
of the UNGC’s impact on business practices (Runhaar and Lafferty,
2009) mainly due to the following reasons: (1) the different
stages of business performance of the participating firms and (2)
the heterogeneous nature of participating firms’ size, geographical
spread, and industry sector. Moreover, previous studies dunder
this third perspectivedhave not investigated whether the firms’
commitment to the UNGC impacts their corporate environmental
performance which is the subject of this paper.
United Nations Global Compact (UNGC) was founded in 2000 as
a response to external institutional pressures and constraints faced
by organizations (Ortas et al., 2015). The initiative serves as a guide
towards greater transparency and accountability in promoting so-
cial responsibility by facilitating learning and discussions through
dialogue among local networks, workshops and partnership pro-
jects that are focused on shaping the business CSR agenda. Also, by
outlining ten principles that relate to human rights, labor, envi-
ronment, and anti-corruption. UNGC is one of the major self-
regulatory global governance initiatives that is crucial in address-
ing negative business practices by setting CSR standards (Abbot and
Snidal, 2010). It has enhanced the understanding of the role and
importance of corporate social responsibility (CSR) and has set
agreeable global business ethical standards. Kohna (2004) credit
the UNGC in the context of shifting strategies for governing
corporate behavior, that is, the shift from hierarchical approaches
to more voluntary and consensual approaches, and from national to
more international governance arrangements. Proponents of the
initiative hold that it promotes dialogue between civil society
groups, private businesses, and international organizations on
global ethical business practices (Rasche, 2011). Critics contend
UNGC members may free-ride and brainwash by joining to create a
positive corporate image and delude on the reality of corporate
responsibility.
Our second point of reflection is from ISO 14001- an EMS
specification of the ISO 14000 standard which looks into five ele-
ments. The five elements include - environmental policy and plan,
implementation and operation of programs, targets, checking and
corrective action, and management review (Babakri et al., 2003).
ISO 14001 is a CSR program that provides a systematic framework
to an organization’s products, services and process to manage
present and future environmental impacts. ISO 14001 encourages
organizations to manage their environmental impacts in a proce-
dural manner. It is based on a series described as the “Deming
Cycle” of plan-do-check-act founded on quality management sys-
tems (Testa et al., 2014). Despite the importance that environ-
mental management system eISO 14001- have in the
3
environmental policy agenda, the debate is still open among re-
searchers and practitioners on their effectiveness in terms of
improvement of environmental performance (Hertin et al., 2008).
Nishitani et al. (2012) and Iwata et al. (2010) concluded that ISO
14001 certified firms are more likely to reduce pollution emission.
Studies by Eng Ann et al. (2006); Fei-Baffoe et al. (2013) and Halis
and Halis (2016) found that EMS adoption to be positively associ-
ated with the environmental performance. Contrary, Hertin et al.
(2008) and Shetty & Kumar (2016) studies found no evidence in
support of organizations certified with EMS have significant posi-
tive impact on environmental performance. A study in Australia by
Tung et al. (2014) concluded that ISO 14001 did not affect envi-
ronmental performance. Zobel (2015) did not observe any statisti-
cally significant difference between ISO 14001 adopters and non-
adopters on environmental performance over a twelve-year.
Gomez and Rodriguez (2011) who compared the toxic release of
ISO 14001 certified and uncertified companies and concluded that
ISO 14001 certification does not result in a reduction of the com-
pany’s environmental polluting footprint. Qi et al. (2014) reported
an unclear correlation between ISO 14001 and environmental
performance. The findings on ISO 14001 are inconclusive, creating
the need for more studies on the effectiveness of the certification
towards environmental performance.
Going by institutional theorists, the survival of an organization
is grounded on its social legitimacy. Thus, it compels it into con-
forming into easily recognizable and acceptable standards (Zucker,
1987). Accreditation with ISO 14001 or UNGC helps firms build-in
structures within the organization that legitimize their internal
and external activities within environmental grounds. By imple-
menting such guidelines, it will make the organization to be
isomorphic to the institutional environment. External constituents
reward the legitimacy with markets, interested employees, con-
tracts and investments, hence surviving. Consequently, normative
pressure and expectations are key in shaping institutional behavior
(Zhang et al., 2013). But, do the adopting firm really meet the
outlined environmental expectations? To shed some understand-
ing on the query, we hypothesize.
1a. Firms accredited with UNGC have more effective waste
management practices.
1b. Firms certified with ISO 14001 have more effective waste
management practices.
In the past studies, regarding the effectiveness of ISO 14001 on
environmental performance perspective, the results mirrored are
univocal. Additionally, similar research to our study by Arimura
et al. (2016) essentially addressed endogeneity in assessing ISO
14001 effectiveness but fell short of raising the situation with
UNGC, a more recent program with broader dimensions. Again,
unlike the Arimura et al. (2016) study, in this research environ-
mental performance is evaluated on a waste management basis.
This study seeks to fill this gap in three folds. First, the study ex-
amines whether ISO 14001 and UNGC- both of which are voluntary
programs-improves firms’ waste management practices conse-
quently improving environmental performance. Advocates of
voluntary environmental programs contend these initiatives pro-
mote beyond compliance environmental protection. A fact attrib-
uted to significant market and regulatory incentives associated
with the adoption of voluntary initiatives (Rivera, 2002). Addi-
tionally, these self-regulatory programs have been reported to
enhance legislation flexibility, cost-effectiveness and promote
technological innovations. However, critics hold that firms partic-
ipate to delude strict regulations and disguise on their poor envi-
ronmental performance (Arora and Cason, 1996). Secondly,
previous studies have considered specific aspects of environmental
4 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636
performance (only for ISO 14001), solid waste generation
(Franchetti, 2011).
3. Research method
3.1. Data and sample
Our findings are based on data collected from a cross-sectional
survey questionnaire within the months of April and June 2019.
There are about 17 million registered firms in Kenya and to
approximate an appropriate sample, we used the National Institute
for Health Research statistics table to approximate an appropriate
sample size. Working with a 95% confidence level, Mathers et al.
(2007) proposed sample size of between 800 and 1000 would be
required as a representative of the population. The geographical
sample frame included Nairobi, Thika, Kiambu and Machakos
counties got through cluster sampling. The three counties account
for about 70% of firms in Kenya cutting across multiple industrial
and commercial sectors. Firms in operation were mapped in streets
within the geographical sample frame from where through random
sampling a final sample was developed.
The questionnaire on environmental performance used in this
study was developed through exhaustive literature review to
identify the thematic areas of the survey. The questions followed a
logical order from general into more specific ones, all which were
clear, short and with instructions. To assess relative measures and
individual statement on various dimensions, the questionnaire
employed Likert and semantic differential scales respectively.
Trained field enumerators were commissioned to interview firm
employees (focal employees) on matters relating to environmental
sustainability of their firms. The focal employees needed to be
knowledgeable and experts (Tatoglu et al., 2019) on corporate social
responsibility matters. However, prior to the actual fieldwork the
draft questionnaire was piloted that led to the identification of
some ambiguity, reordering and pre-code corrections. A total of 18
firms were piloted, a number within the range of 5e50 respondents
as proposed by Mathers et al. (2007). In the actual fieldwork, a total
of 1180 firms were visited. However, only 852 questionnaires were
used for the survey after eliminating 328 responses for which were
micro-enterprises.
The sample included 72% (611 of 852) industrial and 28% (241 of
852) commercial firms. With respect to voluntary environmental
management systems, our focus was on ISO 14001 and UNGC since
they are the most popular, adopted on voluntary basis and applied
across all sectors. Therefore, firms accredited with ISO 14001 were
26% and those certified with UNGC were no more than 11%. VEMS
ratified firms included both the locally owned 72%, foreign owned
12% and those co-owned 16% since organizational proprietorship
influences environmental performance. The data got from the
survey was used to determine if the adoption of VEMS enhanced
corporate environmental performance.
In order to test the reliability of the instruments, internal con-
sistency technique is applied using Cronbach’s Alpha and revealed
that the alpha values (see Table 1) were above the recommended
Table 1
Cronbach’s alpha test.
Variable
Voluntary Environmental Systems
Energy Efficiency
Waste management
Material Management
Institution Pressure
Perceived Financial Performance
threshold of 0.7 (Tatoglu et al., 2019).
3.2. Dependent variable
Waste management was a dummy equal one if a firm adopts all
these practices: waste recycling, wastewater recycling and use of
environmentally safe methods to dispose of waste. Nguyen and
Hens, 2015 advocated for use of waste as an operational measure
of environmental performance. This study uses waste management
as a measure of the effectiveness of ISO 14001 and UNGC because
both help reduces their generation (Arimura et al., 2008). Waste
management is a numeric and non-financial measure of environ-
mental performance indicators that give key information on envi-
ronmental impact (Henri and Journeault, 2008 Jaaffar et al., 2018).
Waste management varies differently in organizations and has
different pressure on the facility. Organizations reduce the gener-
ation of waste even in the absence of regulations and on their
introduction, firms rely on flexible policies such as performance
standards. In doing so, facilities can accrue benefits of reduced
product input costs. Waste recycling, wastewater recycling and use
of environmentally safe disposal methods were measured through
a Likert scale. On a scale of 4-points, respondents were asked the
extent; “The firm undertakes waste recycling”, “The firm un-
dertakes wastewater recycling” and “Disposes of physical waste
through environmentally safe methods”. A binary variable was
used in each of the three proxies. In each case, a firm took the value
of one if a firm undertook waste recycling, wastewater recycling
and uses safe methods to dispose of their waste (2 ¼ to a small
extent, 3 ¼ to a moderate extent, 4 ¼ to a great extent). Otherwise
zero (1 ¼ Not at all).
3.3. Independent variable
UNGC and ISO 14001 are the independent variables that assume
a dummy variable in each case as proposed by Flankova et al.
(2018). If the firm has adopted UNGC ¼ 1 and/or ISO 14001 ¼ 1.
Otherwise ISO 14001 ¼ 0 and/or UNGC ¼ 0.
3.4. Control variables
A study on the performance of EMS in Western Australian
companies by Annandale et al. (2004) identified stakeholder
pressure from the clients, senior management, public and regula-
tors as influencing environmental performance. Further, Arimula
et al. (2008) observed that primary consumers desire of a com-
pany to have a green image and firm produce export destinations to
sway environmental performance. Regulatory pressure both legis-
lative and political will were found to influence organization
environmental performance. We control for stakeholders’ pressure
from regulators, major competitors, community, media and
possible influences from the management through a reward based
on environmental performance.
Multinational enterprises have a higher likelihood of adopting
voluntary environmental programs to reduce risks on their
Number of Items Cronbach’s Alpha Coefficient
8 0.7580
10 0.8952
7 0.8591
8 0.8536
21 0.9134
7 0.9288
 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636 5

foreignness. Firms exporting to foreign markets are likely to have Pr(WASTEMGTi ¼ 1|EMSi,X1i) ¼ F (dEMSi þ x0 1ib1), (3)
more pressure to adopt international certification programs
(Christmann and Taylor, 2006). Arimula et al. (2008) controlled for The distribution function of the standard normal random vari-
various firm characteristics the number of employees as a proxy for able is represented as F. Average partial effect in the interest of
firm size, age of the facility and listing with the stocks exchange. concern in EMS.
Welch and Hibiki (2003) controlled firm financial performance and
reported there wasn’t any relationship, Unlike McGuire (2010) who EX1[Pr(WASTEMGT ¼ 1|EMS ¼ 1, x1) e Pr(WASTEMGT ¼ 1| EMS ¼ 0,
found out that firms with higher total revenue were likely to adopt x1)], (4)
ISO 14001. As a result, we have controlled for various firm charac-
teristics among them firm size (proxied as the number of em- Alternatively approximated as;
ployees), ownership, market, industrial sector and organizational
strategies such as sustainability department and environmental X
N     
policy. 1∕N F d þ x’1 b1  F x’1 1b1 ; (5)
i¼1
There are equivocal findings on the effect of EMS on environ-
mental waste in the extant literature. Reasons that Arimura et al.
(2016) argue to be brought up by the difference in regulatory set-
tings among different firms and the concerns of endogeneity in the
studies that may be unobservable yet positively related to envi- 3.7. Endogeneity of EMS (ISO 14001 and UNGC)
ronmental performance. In addition, there are different measures
of environmental performance that range from specific indicators We now assume that EMS (ISO 14001 and UNGC) is an endog-
such as air pollution to multiple natural resources measures. enous variable. A bivariate probit model will allow us to identify the
potential endogeneity of EMS.
3.5. Model specification
EMSi ¼ 1[EMS*i  0 (6)
Assuming no endogeneity of ISO 14001 and UNGC, we use a
simple probit model to estimate their effects on environmental EMS*i ¼ X0 2ib2þ ε2i (7)
performance. Our data set lacks any traditional instrument and
newly proposed ones did not fulfill all the validity requirements. Where i ¼ ISO 14001, UNGC.
Hence we rely on Altonji et al.’s (2005) approach for lacking in- Also, EMS*- latent variable-represents the positive influences of
struments. The method assumes that the degree of selection of having an EMS while X0 2 is a group of control variables and ε, the
unobservables is equal to the degree of selecting the observables. error term. The assumption in this model is that the normal dis-
Either of these components of environmental performance may or tribution of ε is bivariate.
may not have influences on EMS (ISO 14001 and UNGC) adoption. We have no credible instruments and we estimate the bivariate
Hence the degree of correlation between unobservable and probit model with X1 ¼ X2 ¼ X. Therefore, we opt for an assumption
observable elements of performance is the same. Generally, it is that replicates results derived via the common standard instru-
expected that the level of the organization’s waste management mental variable approach. Altonji et al. (2005) brought up the idea
practices will improve provided that the net benefit from waste of using the selection of observables to study, the bias due to un-
management initiatives can be demonstrated. Waste management observables and showed how such information is incorporated in
improvement is dependent on several factors including the adop- estimation. The adoption of EMS takes the following model;
tion of EMS (ISO 14001 and UNGC). The resulting model takes the Let WASTEMGT* be determined by
form;
WASTEMGT* ¼ dEMS þ w0 g (8)
WASTEMGTi ¼ 1[WASTEMGT*i  0] (1)
w is the vector with both observed and unobserved variables that
WASTEMGT*i ¼ dEMSi þ X1ib1 þ ε1i (2) determines WASTEMNGT*, w¼ (w1, K, wk)’ and g is the coefficient
vector. g ¼ (g1, K, gk)0 . We further define dummy variable sj indi-
Where (1) is an indicator function. WASTEMNGT*i is a latent vari- cating if wj is observed. In both equations, observed and unob-
able that represents the net value got from practicing waste man- served variables are given by
agement. EMSi assumes the value of one of the facilities was
certified with UNGC or/and ISO 14001. It is a dummy variable with X
k X
k

d coefficient. Further, X1i and b1 represent the vector of the study x’gx ¼ sjwj; v ¼ ð1  sjÞwjgj; (9)
j¼1 j¼i
control variables, corresponding coefficients respectively and
idiosyncratic error term respectively.
In the above equation, x represents the observable portion of w
In this case, if an organization undertook waste management,
with coefficients contained in gx (a subvector of g) and v is an error
then WASTEMGT ¼ 1.
component that represents unobserved variables in w. Considering
x0 gx and v, we can rewrite equation (8) as;
3.6. Standard probit model
WASTEMNGT* ¼ dEMSþ x0 gX þ v, (10)
Let us assume that EMS is exogenous. Hence, they are inde-
pendent of the error term. Also, we assume that the error term is In this equation, it is highly unlikely that the observed control
standard normally distributed. These assumptions result in a probit variables x are uncorrelated with v.
model Equation (10) can be rewritten as

WASTEMGTi ¼ 1; WASTEMNGT* ¼ dEMSþ x0 b1þε1, (11)

6 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636

Here, b1 is defined to certify COV(ε1, X) ¼(i.e., x is exogenous in
equation; b1 therefore accounts for both the direct effect of x on
WASTEMNGT (i.e., gx) and the interaction between x and v. Now,
we assume that the linear projection of EMS on x0 b and ε1 is pre-
sented as:
Proj(EMS*|X0 b1, ε1) ¼ 40þ 4X0 b1X0 b1þ4ε1ε1
4X0 b1(þ4ε1) shows the influence of the observable (unobserv-
able) component of the WASTEMNGT results on EMS adoption
respectively. It represents a selection on observables (selection on
unobservable). The term 4ε represents the two assumptions made
by Altonji et al. (2005) on the extreme conditions about the degree
of selection on observable. The conditions provide the identifying
information for the upper and lower bound estimates of EMS ef-
fects d. The first extreme condition is that the part of WASTEMNGT
related to the unobservables has no relationship with EMS*, which
is 4ε1 ¼ 0. If this condition holds, the conclusion is no endogeneity
of EMS.
The second extreme condition is when a section of WASTEMNGT
related to the observables has the same relationship with EMS* as
that related to the unobservables, i.e., 4ε1 ¼ 4x0 b1. Considering data
collection restrictions, we interpret the elements of X to be a
random subset of several variables w. It agrees to the assumption
that each Sj in the equation is independently and identically
distributed binary random variable equal to one with the same
probability for all j. When w gets large, the second condition be-
comes approximately true for all projections of sj. The two extreme
conditions limit the degree of selection on unobservables between
zero and an upper value that is equal to the degree of selection on
the observables.
Each extreme represents a different identifying assumption for
the d coefficient. The first extreme condition corresponds to the
limit that r ¼ 0 in the bivariate probit condition described as
x1 ¼ x2 ¼ x. To obtain the upper-value approximation of thε d co-
efficient we estimate a univariate probit model for performance.
The second condition can be shown to be approximately equal to
the following upper limit for r
COVðxb1; xb2Þ
r¼
VARðX b1Þ
Extending the log estimate over (d, b1, b2) subject to the limits on
r given above provides a lower limit approximation of the
d coefficient.
unknown and are further estimated with b value. Consequently, the
approximations of various observed values of y, may be interpreted
as the intercepts like in (i) through the threshold parameters.
In the survey questionnaire, responses are an expression one’s
feelings and are mutually dependent on measurable and unob-
servable factors designated as ε. The questionnaire ordinal scale is
(12) from one, meaning ‘not at all’ to four ‘implying to a great extent’ all
that represent varied feelings. The responses are assumed to denote
scores that closely related to the attitude on the given question.
With the assumption that εis normally distributed with an
approximated zero value and a variance of one, then we can have;
Prob(Y ¼ 0|X) ¼ 4(-x’b)
Prob(Y ¼ 1|X) ¼ 4(m1 - x’b) - 4(x’b)
Prob(Y ¼ 2|X) ¼ 4 (m2 - x’b) - 4(m1 - x’b)
Prob(Y ¼ j|X) ¼ 1 - 4(mj-1 - x’b) (2.3)
In order to have all the probabilities as positive, we need to have,
0 < m1< m2 < … < mj-1 (2.4)
As the conventional practice, the probabilities of the indepen-
dent variables of x should not be equivalent with those of the co-
efficient on the marginal effects. As an example of a scenario with
five categories, the model will incorporate three unknown
threshold parameters.
Considering a case with five categories, the model will have
three unknown thresholds with the first probability normalized to
zero and others as;
Prob(Y ¼ 0|X) ¼ 4(-x’b)
Prob(Y ¼ 1|X) ¼ 4(m1 - x’b) - 4(- x’b)
Prob(Y ¼ 2|X) ¼ 4 (m2 - x’b) - 4(m1 - x’b)
Prob(Y ¼ 3|X) ¼ 4 (m3 - x’b) - 4(m2 - x’b)
(13)
Prob(Y ¼ 4|X) ¼ 1 (2.5)
Subsequently, the matching marginal effects of the changes in
independent variables for the above probabilities are:

v Probðy ¼ 0jXÞ
3.8. Robustness check: ordered probit ¼ 4ðx’ bÞb
vX

To carry out the robustness check, we follow the proposal by

Alauddin and Tisdell (2006), by developing the model below that is vProbðy ¼ 1jXÞ
¼ ½4ðX ’ bÞ  4ðm1  x’ bÞb
built around a latent regression. vX

y* ¼ xb þ a (2.1) 
vProbðy ¼ 2jXÞ
¼ 4ðm1  x’ bÞ  4ðm2  x’ bÞ b
In the illustrated case, y* is unobserved while the observed is: vX

Y ¼ 0 if y*  0
vProbðy ¼ 3jXÞ
¼ ½4ðm2  x’ bÞ  4ðm3  x’ bÞb
Y ¼ 1 if 0 < y*  m1 vX

Y ¼ 2 if m1< y*  m2
vProbðy ¼ 4jXÞy
¼ 4ðm3  x’ bÞ (2.6)
Y ¼ j if y*  mj-1 (2.2)
vX
The analytical model described above, was used to check the
The values; m1, m2, and mj-1 signify threshold parameters that are robustness of our findings (see Table 6).
 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636 7

4. Results likely to file a lawsuit against the firms due to environmental

Table 2 presents the descriptive statistics for the study variables
and the control variables of the sample. The findings are discussed
below.
Variables of Interest. The results indicated 44% of the firms
undertook waste management through the three methods-waste,
wastewater recycling, and safe methods disposal. Additionally,
regarding the EMS, 26% and 12% of the sampled firms were ISO
14001 and UNGC certified respectively.
Firm Characteristics. A proportion of 78% of the firms based
their investment decisions on whether the proposed investment
would have effect on the reduction on waste and pollution while
48% rewarded their senior management based on environmental
performance. Further, 69% of the firms had an environmental policy
while 34% reported having a sustainability department. Moreover,
29% of the firms had women in Top Management Team (TMT). The
average years of experience of women in TMT was 7.22 years with a
standard deviation of 6.90 and a maximum of 50 years. Subse-
quently, the average number of employees was approximately 247
employees with a standard deviation of 712.49, a minimum of 10
employees and a maximum of 7300 employees. Further, the
average age of the business was 21.23 years with a standard devi-
ation of 20.47 years, a minimum of 0.42 years and a maximum of
123 years. Nevertheless, 72% and 12% of the sampled firms were
Kenya fully owned and foreign fully owned respectively. Finally,
40% of the sampled firms were environmentally sensitive
industries.
Stakeholders’ Pressure. A proportion of 86% of the firms indi-
cated that regulators tried to initiate environmental activities while
85% indicated that major customers frequently required firms to
adopt environmentally responsible businesses. Additionally, 63% of
the firms reported that community environmental groups were
misconduct while 31% revealed that the media closely monitored
their activities.
Financial Performance. The findings revealed that 53% and 59%
of the firm’s perceived returns on assets and sales respectively were
better or much better in the last financial year. A proportion of 36%
of the sampled firms exported their products.
Table 3 presents the correlations between the study and control
variables. The dependent variable-waste management is positively
and significantly correlated with ISO 14001 adoption, UNGC
adoption, investment decisions that results to reduced waste and
pollution. Additionally, they are positively and significantly related
to the rewarding of senior management based on environmental
performance, environmental policy, stakeholder’s holders’ pressure
e regulators, major customers, competitors, community groups
and the media. Further, the number of employees and the presence
of the sustainability department are positively and significantly
associated with waste management. The perceived financial per-
formance indicators-return on assets and sales had no significant
correlation with waste management.
Our hypothesis (1a) purports that firms accredited with UNGC
have more effective waste management practices. Model 2 from
Table 4 indicates that the bivariate waste management marginal
effects for UNGC are positive and significant. Thus, the probability
to improve environmental performance by adopting the three
waste management proxies increases by 34.0 percentage points if
the firm is a participant of UNGC. The results support affirms our
hypothesis (1a).
Our hypothesis (1b) predicts that firms that have adopted ISO
14001 certification have improved waste management. Model 4
from Table 5 reveals that the bivariate waste management marginal
effects for ISO 14001 adoption are positive and significant. There-
fore, the probability to improve environmental performance by

Table 2
Descriptive statistics.

Variable Description of Variable N Mean Standard Minimum Maximum

Name Deviation

Variables of Interest
Wastemgt The firm undertakes waste management, wastewater recycling, and safe waste disposal dummy ¼ 1 if 852 0.44 0 1
yes zero otherwise
ISO14001 The firm is ISO 14001certified yes ¼ 1 852 0.26 0 1
UNGC The firm Participant of UN Global Compact Sustainability Initiative yes ¼ 1 852 0.12 0 1
Firm characteristics
investdecision Investment (R&D, processes, products, machinery, etc.) decisions are also based on reduced waste 852 0.78 0 1
and pollution, Not at all ¼ 0
Smgtreward Senior management is rewarded based on the environmental performance of the firm yes ¼ 1 852 0.48 0 1
Environpolicy The firm has an environmental policy yes ¼ 1 852 0.69 0 1
sharewomenTMT Share of women in the top management team 852 0.29 0 1
womenexpTMT Average years of experience of women in top management team 852 7.22 6.90 0 50
employees Number of Employees 852 247.23 712.49 10 7300
Firmage Age of the Business in Years 852 21.23 20.47 0.42 123
Export Firm Exports the products 852 0.36 0 1
sustaindept The firm has sustainability department yes ¼ 1 852 0.34 0 1
Kenyan Kenyan fully owned 852 0.72 0 1
Foreign Foreign fully owned 852 0.12 0 1
ESI Environmental sensitive industries 852 0.40 0 1
Stakeholders’ Pressure
regulatorinit Regulators try to initiate environmental activities of companies, Not at all ¼ 0 852 0.86 0 1
majorcustomers My firm’s major customers frequently require my firm to adopt environmentally responsible 852 0.85 0 1
business, Not at all ¼ 0
competitors Our strongest competitors take a leading role in environmental responsibility, Not at all ¼ 0 852 0.66 0 1
Community Community environmental groups are likely to file a lawsuit against our firm in the event of an 852 0.63 0 1
environmental misconduct, Not at all ¼ 0
Media Our firm’s activities are closely monitored by the media, Not at all ¼ 0 852 0.31 0 1
Financial Performance
rosgrowth perceived Return on sales ¼ 1 if better or much better in the last financial year. 852 0.59 0 1
roagrowth perceived Return on assets ¼ 1 if better or much better in the last financial year. 852 0.53 0 1

This table presents descriptive statistics. The sample includes 852 firms in Kenya.
8 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636

22
adopting the three waste management proxies increases by 31.0
percentage points if the firm is ISO 14001 certified. The results

0.02 1
support our hypothesis (1b) (see Table 5).
21

The findings from Table 4 model 2 indicates that the bivariate

0.02 0.65*** 1
waste management marginal effects for a firm’s investment de-

0.02 0.02
cisions are based on waste reduction and pollution, senior man-
20

This table presents pairwise correlations among dependent and independent variables. The sample includes 852 firms in Kenya. *, ** and *** denote significance at the 10%, 5% and 1% levels respectively.
0.02 1
agement rewarded based on the environmental performance and
the number of employees are positive and significant. Therefore,
19

1
the probability for a firm to improve waste management increases

0.58***

0.09***
by 17.0 percentage points if the firm aligns investment decisions

0.04

0.03
with waste reduction and pollution. Additionally, the probability
18

1
that a firm improves waste management increases by 15.0 per-

0.09**
0.03
0.07*
centage points if the firms incorporate environmental performance

0.00
0.01
17

1
when rewarding senior management. Moreover, the probability to

0.12***
0.17*** improve waste management increases by 0.03 due to a 100% in-

0.11***
0.12***
0.01
crease in the number of employees. Further, the bivariate waste
0.05
16

management marginal effects for the sustainability department

0.15***
0.15***

0.10***
0.13***
0.06*

and environmentally sensitive industries are positive and signifi-

0.00

0.02 cant. Therefore, the probability that a firm improves waste man-
15

agement increases by 6.0 percentage points if a firm has a

0.15***

0.10***
0.49***
0.20***
0.20***

0.14***
0.14***

sustainability department. Additionally, the probability to improve

0.01

waste management increases by 8.0 percentage points if the firm is

14

environmentally sensitive. According to model 2 from Table 4,

0.09***
0.24***
0.19***
0.09***

0.14***
0.14***

bivariate waste management marginal effects for stakeholders’

0.07**
0.03
0.05

pressure-regulators and community groups are positive and sig-

13

nificant. This implies that the probability to improve waste man-

0.10***
0.62***

0.06*

agement increases by 11.0 percentage points due to pressure from

0.03

0.02
0.02
0.07*
0.02

0.05
0.04

regulators while increases by 14.0 percentage points due to pres-

12

sure from environmental community groups.

0.12***
0.30***
0.13***
0.13***
0.29***

0.09***
0.06

0.05
0.04

0.00

0.12
11

4.1. Robustness test

1
0.24***

0.10***
0.17***
0.08**

0.02

0.02
0.06*
0.03
0.05

0.00

0.05
0.05

From Table 6, the marginal effects for UNGC are significant.

10

Therefore, our results are robust. Hence, regardless of the model

0.07**
0.25***
0.21***

0.15***
0.08**

choice and specification, UNGC is found to influence waste man-

0.03
0.05
0.02
0.04
0.04

0.00

0.02

0.01

agement. This supports our findings.

9

1
0.37***
0.23***
0.16***

0.11***
0.17***
0.01

0.01
0.06*

5. Discussion
0.05
0.05

0.02

0.00
0.02
0.05
8

1
0.26***
0.28***
0.19***
0.20***

0.17***
0.11***
0.10***
0.13***

0.06*

This research contributes to the growing debate on voluntary

0.02
0.05

0.02

0.05
0.02
0.05

environmental programs that advocate for sustainable production

7

through resource management and sound waste management.

0.08**
0.31***
0.28***
0.14***
0.18***
0.17***

0.19***

0.21***
0.23***

0.14***
0.08**

0.09**

0.09**
0.00

Adoption of ISO 14001 and UNGC are viewed as environmental

0.06*
0.02

management actions focused on improving energy and water

6

management, while mitigating environmental impacts from waste.

0.30***
0.19***
0.29***
0.26***
0.22***
0.32***

0.17***

0.14***
0.28***
0.07**

0.03
0.02
0.06

0.03

0.03
0.05
0.02

Past studies have examined the use of natural resources or singled

5

out various measures of environmental performance in their

0.08**
0.071**
0.28***
0.28***
0.21***
0.18***
0.20***
0.20***
0.10***

0.20***

assessment of voluntary program-ISO 14001-effectiveness. How-

0.08**

0.08**

0.01
- 0.00
- 0.01

0.03

0.01

0.05

ever, the findings have been somewhat mixed and therefore we aim
4

to join this conversation to provide more clarity.

0.12***

0.12***
0.15***
0.14***

0.10***

0.11***
0.25***

0.10***
0.29***
0.20***
0.19***
0.21***

0.14***

Thus, our study provides an important perspective on the role of

0.07**

0.07**

0.07**
0.06*

0.06*
0.06

ISO 14001 and UNGC as voluntary strategies from a waste man-

3

agement approach. Additionally, our results are based on a devel-

0.10***
0.21***
0.17***
0.24***
0.24***
0.13***
0.16***
0.13***

0.18***

0.19***
0.12***
0.12***
0.22***

0.09***

oping economy that is responding to environmental challenge

0.07**

0.08**
0.06*
0.03

0.01

0.03

through coordinated efforts of the National, County governments

2

and the private sector especially with the requirements of the na-
0.07**
0.18***
0.15***
0.27***
0.36***
0.24***
0.20***
0.20***
0.16***
0.27***
0.21***

0.22***
0.11***
0.15***
0.24***
0.07**

tional government to fulfill the requirement of the Paris Agreement.

0.03

0.00
0.00
0.02
0.05

In the efforts to foster climate change adaptation and resilience, the

1

Kenyan government enacted the Climate Change Act in 2016.

13.lnwomenexpTMT
12.sharewomenTMT
8.majorcustomers

Subsequently, in 2018, a five-year National Climate Change Action

4.investdecision

14.lnemployees
6.Environpolicy

10.Community

17.sustaindept
7.regulatorinit
5.Smgtreward

Plan (NCCAP) was created as a climate change action strategy

9.competitors

21.roagrowth
20.rosgrowth
15.lnfirmage
1.wastemgt
2.ISO14001
Correlations.

(Government of Kenya, 2018). The action plan underpins water,

18.kenyan
19.foreign
16.export
11.media
3.UNGC

sanitation, and waste as areas of interest, that have shaped the

22.ESI
Table 3

scholarship discussion in this paper. NCCAP is in alignment with

sustainable development goals as well as the Government of Kenya
 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636 9

Table 4
Effects of UNGC on waste management.

Model 1 (DP Var: waste Model 2 (DP Var: waste management & UNGC)
management)

Probit Probit Marginal Bivariate wastemgt Bivariate UNGC Bivariate wastemgt Marginal Bivariate Bivariate Joint Marginal
Coefficient Effects coefficient coefficient Effects UNGC Effects
Marginal
Effects

UNGC 0.16 (0.156) 0.06 (0.062) 1.10** (0.504) 0.34** (0.152)

investdecision 0.56*** 0.21*** (0.045) 0.55*** (0.134) 0.04 (0.180) 0.17*** (0.040) 0.01 0.01 (0.009)
(0.134) (0.027)
Smgtreward 0.54*** 0.21*** (0.039) 0.50*** (0.107) 0.08 (0.150) 0.15*** (0.032) 0.01 (0.022) 0.02* (0.009)
(0.104)
Environpolicy 0.13 (0.116) 0.05 (0.044) 0.11 (0.115) 0.27 (0.174) 0.03 (0.035) 0.04 (0.026) 0.02** (0.009)
regulatorinit 0.37** 0.14** (0.058) 0.37** (0.163) 0.04 (0.246) 0.11** (0.050) 0.01 (0.037) 0.01 (0.011)
(0.166)
majorcustomers 0.29** 0.11* (0.058) 0.26 (0.159) 0.26 (0.251) 0.08 (0.049) 0.04 (0.037) 0.12* (0.010)
(0.161)
competitors 0.08 (0.114) 0.03 (0.045) 0.07 (0.111) 0.12 (0.159) 0.02 (0.034) 0.02 0.01 (0.010)
(0.024)
Community 0.47*** 0.18*** (0.038) 0.44*** (0.106) 0.11 (0.154) 0.14*** (0.032) 0.02 (0.023) 0.02** (0.008)
(0.105)
media 0.004 0.001 (0.044) 0.09 (0.120) 0.42*** (0.149) 0.03 (0.037) 0.06*** 0.03** (0.012)
(0.114) (0.022)
sharewomenTMT 0.14 (0.280) 0.05 (0.109) 0.02* (0.280) 0.89*** (0.346) 0.01 (0.086) 0.13*** 0.05** (0.023)
(0.052)
lnwomenexptmt 0.02 (0.062) 0.01 (0.024) 0.03 (0.060) 0.13 (0.082) 0.01 (0.018) 0.02 0.01 (0.005)
(0.012)
lnemployees 0.15*** 0.06*** (0.017) 0.11** (0.050) 0.16*** (0.052) 0.03** (0.016) 0.02*** 0.01*** (0.004)
(0.044) (0.008)
lnfirmage 0.02 (0.053) 0.01 (0.021) 0.05 (0.055) 0.21*** (0.081) 0.01 (0.017) 0.03*** 0.01** (0.005)
(0.012)
export 0.11 (0.103) 0.04 (0.040) 0.05 (0.108) 0.44*** (0.135) 0.01 (0.033) 0.07*** 0.03*** (0.011)
(0.020)
sustaindept 0.27** 0.10** (0.042) 0.20* (0.115) 0.29** (0.140) 0.06* (0.035) 0.04** 0.03** (0.011)
(0.108) (0.021)
kenyan 0.21 (0.136) 0.01 (0.053) 0.17 (0.139) 0.19 (0.163) 0.05 (0.043) 0.03 0.02 (0.012)
(0.024)
foreign 0.12 (0.181) 0.05 (0.069) 0.13 (0.182) 0.20 (0.216) 0.04 (0.056) 0.03 (0.032) 0.01 (0.015)
rosgrowth 0.13 (0.130) 0.05 (0.051) 0.10 (0.131) 0.09 (0.154) 0.03 (0.040) 0.01 0.01 (0.010)
(0.023)
roagrowth 0.06 (0.131) 0.02 (0.051) 0.11 (0.131) 0.26* (0.151) 0.03 (0.040) 0.04* (0.023) 0.01 (0.010)
ESI 0.21** 0.08** (0.039) 0.26*** (0.098) 0.41*** (0.143) 0.08*** (0.030) 0.06*** 0.02** (0.008)
(0.100) (0.021)
Constant 2.43*** 2.18*** (0.350) 3.42*** (0.404)
(0.314)
No. of 852 852 852 852
Observation
Wald c2(20) 194.72***
Wald c2(39) 405.90***
rho 0.54 (0.273)
Wald test of c2(1) ¼ 2.440
rho ¼ 0:

Robust standard errors in parenthesis accounts for heteroscedasticity. *, ** and *** denote significance at the 10%, 5% and 1% levels respectively.

Big four agenda.
Kenya is gearing up towards a circular economy that is capable
of diverting 90% of collected waste away from disposal sites
(Government of Kenya, 2018). Based on our conclusions, firms that
are UNGC certified are more likely to recycle their waste. A factor
that may be attributed to the sharing of emerging best industrial
practices that offer practical solutions and approaches to waste
management. In response to growing water scarcity, the action plan
intends to increase annual per capita water availability by devel-
oping the water infrastructure that incorporates; reducing waste,
reusing, and recycling. According to the plan, firms in Kenya are
adopting resource efficiency methods such as water efficiency
planning and water audits. This is in line with our findings that
firms that have adopted UNGC have a higher probability to recycle
their wastewater. Again, the program may be interpreted to be
allowing adopters to access extensive knowledge and experience
with green growth issues through extensive partnership and
networking. UNGC advances the idea of investment in new tech-
nologies and the adoption of quality management systems. Varia-
tions in the development and implementation of both ISO 14001
and UNGC significantly influence the outcomes of environmental
performance. ISO 14001 and UNGC are similar and are advocating
for continuous improvement, conversely, ISO 14001 is seen to focus
on the process, unlike UNGC that is specific on results. Thus, the
adoption of ISO 14001 may not guarantee observable increase in
scores on earmarked environmental performance indicators such
as our study proxies of waste management.
Our findings agree with those reported by Henriques and
Sadorsky (2007) that regulators play a leading role in advocating
for environmentally responsible practices. Hence, ISO 14001 certi-
fication may be symbolic among Kenyan firms partly in the aspi-
ration for legitimacy ‘ritual integration’. As a reverberation of the
ideals of institutional theory that conforms to environmental rules,
our results show that firms attempt to allocate resources,
10 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636

Table 5
Effects of ISO14001 on waste management.

Model 3 (DP Var: waste Model 4 (DP Var: waste management & ISO14001)
management)

Probit Probit Marginal Bivariate wastemgt Bivariate ISO14001 Bivariate wastemgt Bivariate Bivariate Joint Marginal
Coefficient Effects coefficient coefficient Marginal Effects ISO14001 Effects
Marginal
Effects

ISO14001 0.13 (0.111) 0.05 (0.044) 1.05** (0.460) 0.31** (0.124)

investdecision 0.55*** 0.20*** (0.045) 0.45*** (0.143) 0.41*** (0.147) 0.13*** (0.044) 0.11*** 0.08*** (0.015)
(0.134) (0.041)
Smgtreward 0.52*** 0.20*** (0.040) 0.39*** (0.137) 0.34*** (0.112) 0.12*** (0.043) 0.09*** 0.09*** (0.019)
(0.105) (0.031)
Environpolicy 0.12 (0.117) 0.04 (0.045) 0.01 (0.124) 0.45*** (0.133) 0.003 (0.037) 0.12*** 0.06*** (0.017)
(0.036)
regulatorinit 0.36** 0.14** (0.058) 0.32** (0.160) 0.13 (0.186) 0.10** (0.048) 0.04 (0.051) 0.05** (0.020)
(0.166)
majorcustomers 0.29* (0.162) 0.11* (0.058) 0.22 (0.159) 0.33* (0.181) 0.06 (0.048) 0.09* (0.050) 0.06*** (0.019)
competitors 0.08 0.03 (0.045) 0.11 (0.114) 0.12 (0.124) 0.03 (0.034) 0.03 (0.034) 0.01 (0.018)
(0.114)
Community 0.49*** 0.18*** (0.039) 0.53*** (0.101) 0.35*** (0.111) 0.16*** (0.030) 0.10*** 0.00 (0.017)
(0.106) (0.030)
media 0.01 (0.114) 0.002 (0.045) 0.04 (0.114) 0.16 (0.116) 0.01 (0.034) 0.05 (0.032) 0.02 (0.019)
sharewomenTMT 0.14 (0.279) 0.06 (0.109) 0.03 (0.274) 0.51* (0.281) 0.01 (0.081) 0.14* (0.078) 0.07 (0.047)
lnwomenexptmt 0.01 (0.061) 0.01 (0.024) 0.02 (0.059) 0.03 (0.063) 0.01 (0.018) 0.01 (0.017) 0.002 (0.010)
lnemployees 0.15*** 0.06*** (0.017) 0.12** (0.048) 0.08* (0.044) 0.03** (0.015) 0.02* (0.012) 0.02*** (0.007)
(0.044)
lnfirmage 0.02 0.01 (0.020) 0.04 (0.051) 0.09* (0.054) 0.01 (0.015) 0.02* (0.015) 0.01 (0.009)
(0.052)
export 0.12 (0.102) 0.04 (0.040) 0.10 (0.101) 0.02 (0.109) 0.03 (0.030) 0.004 (0.030) 0.12 (0.017)
sustaindept 0.26** 0.10** (0.042) 0.15 (0.126) 0.28** (0.109) 0.05 (0.038) 0.08*** 0.06*** (0.020)
(0.108) (0.030)
kenyan 0.20 0.08 (0.053) 0.10 (0.143) 0.28** (0.132) 0.03 (0.043) 0.08** 0.06** (0.026)
(0.136) (0.036)
foreign 0.11 0.04 (0.069) 0.06 (0.174) 0.15 (0.181) 0.02 (0.052) 0.04 (0.050) 0.03 (0.025)
(0.180)
rosgrowth 0.14 0.06 (0.051) 0.18 (0.128) 0.18 (0.133) 0.05 (0.038) 0.05 (0.037) 0.01 (0.020)
(0.131)
roagrowth 0.05 0.02 (0.051) 0.04 (0.128) 0.03 (0.130) 0.01 (0.038) 0.01 (0.036) 0.01 (0.020)
(0.131)
ESI 0.20** 0.08** (0.039) 0.16 (0.100) 0.12 (0.102) 0.05 (0.030) 0.03 (0.028) 0.03* (0.017)
(0.099)
Constant 2.14*** (0.362) 2.57*** (0.320)
No. of 852 852 852 852
Observations
Wald c2(20) 195.98***
Wald c2(39) 476.48***
rho 0.56 (0.273)
Wald test of c2(1) ¼ 2.517
rho ¼ 0:

Robust standard errors in parenthesis accounts for heteroscedasticity. *, ** and *** denote significance at the 10%, 5% and 1% levels respectively.

customers, community and employees appropriately. That means
that firms that are perceived to be better in terms of environmental
performance are gaining the benefits. Such motivation is helping to
influence the manager’s attitudes, intentions, subjective norms and
behavioral control towards the implementation of EMS.
6. Conclusion
We acknowledge our results are inconclusive in some respects.
First, the validity of our findings partly hinges on the acceptance of
our independent variables as measures of EMS efficacy. Thus, we
call for research confirmation, extension and elaboration using
internal or government sponsored EMS. Secondly, firm approaches
such as rewarding the senior management based on environmental
performance merits and the presence of a sustainability depart-
ment show significant influence on environmental performance.
Our research fails to identify if such strategies were guidelines
within ISO 14001 and UNGC, legal requirements or the firm’s own
initiative. Future studies can explore these variables in addition to
specific climate-resilient technologies that have been adopted in
waste management and their impact on corporate environmental
performance. Possibly some of the NCCAP provisions of training,
financing green startups and creation of awareness could have
enhanced environmental performance. We did not explore this
alternative, but future scholars may do so. Both ISO 14001 and
UNGC are third party programs, a research into the efficacy of in-
dustrial and self-monitoring voluntary programs can further
expound our understanding of VEMS in emerging economies.
Nonetheless, our study is not short on offering insights from those
fruitful perspectives.
Practically, the government should subsidize VEMS imple-
mentation to enable access to external expertise crucial in
enhancing environmental management. Moreover, to promote
voluntary culture in environmental protection among firms, the
government opts to develop fiscal and tax policies as incentives.
This should be in line with enabling access to any sort of technical
assistance. Again, with the observed influence from both the reg-
ulators and communities and the insignificance of other
 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636 11

Table 6
Ordered Probit Model: Dependent Variable: Waste management.

Coefficient Marginal Effects

Not at all (1) To a small Extent (2) To a moderate Extent (3) To great Extent (4)

ISO14001 0.11 (0.093) 0.02 (0.015) 0.02 (0.020) 0.03 (0.024) 0.01 (0.012)
UNGC 0.25* (0.130) 0.04** (0.018) 0.06* (0.033) 0.06** (0.031) 0.03* (0.020)
investdecision 0.68*** (0.105) 0.15*** (0.029) 0.08*** (0.013) 0.17*** (0.025) 0.06*** (0.009)
Smgtreward 0.31*** (0.089) 0.06*** (0.016) 0.06*** (0.018) 0.08*** (0.023) 0.04*** (0.011)
Environpolicy 0.43*** (0.097) 0.09*** (0.022) 0.07*** (0.014) 0.11*** (0.025) 0.04*** (0.010)
regulatorinit 0.30** (0.125) 0.06** (0.029) 0.05*** (0.014) 0.08** (0.032) 0.03*** (0.011)
majorcustomers 0.24* (0.124) 0.05* (0.027) 0.04** (0.016) 0.06* (0.032) 0.02** (0.011)
competitors 0.08 (0.092) 0.01 (0.016) 0.02 (0.019) 0.02 (0.023) 0.01 (0.011)
Community 0.29*** (0.086) 0.06*** (0.017) 0.05*** (0.015) 0.07*** (0.022) 0.03*** (0.010)
media 0.12 (0.094) 0.02 (0.018) 0.02 (0.017) 0.03 (0.024) 0.01 (0.010)
sharewomenTMT 0.31 (0.228) 0.05 (0.041) 0.06 (0.045) 0.08 (0.058) 0.04 (0.027)
lnwomenexptmt 0.04 (0.050) 0.01 (0.009) 0.01 (0.010) 0.01 (0.013) 0.004 (0.006
lnemployees 0.06* (0.036) 0.01* (0.007) 0.01 (0.007) 0.02* (0.009) 0.01* (0.004)
lnfirmage 0.001 (0.042) 0.00 (0.007) 0.00 (0.008) 0.00 (0.011) 0.00 (0.005)
Export 0.21** (0.084) 0.04** (0.014) 0.04** (0.018) 0.05** (0.021) 0.03** (0.011)
sustaindept 0.20** (0.088) 0.04** (0.015) 0.04** (0.019) 0.05** (0.022) 0.03** (0.012)
kenyan 0.15 (0.106) 0.03 (0.018) 0.03 (0.023) 0.04 (0.027) 0.02 (0.014)
foreign 0.10 (0.147) 0.02 (0.029) 0.02 (0.024) 0.03 (0.038) 0.01 (0.015)
rosgrowth 0.09 (0.102) 0.02 (0.019) 0.02 (0.019) 0.02 (0.026) 0.01 (0.012)
roagrowth 0.02 (0.102) 0.004 (0.018) 0.005 (0.020) 0.01 (0.026) 0.003 (0.012)
ESI 0.10 (0.080) 0.02 (0.014) 0.02 (0.019) 0.02 (0.020) 0.01 (0.010)
N 852
LR c2 (21) 293.16***

The table presents ordered Probit model of waste management. Robust standard errors in parenthesis account for heteroscedasticity. *, ** and *** denote significance at the
10%, 5% and 1% levels respectively.

stakeholders, there is an opportunity to introduce government Declaration of competing interest

sponsored voluntary programs. The programs offer greater di-
versity for stakeholder involvement that are critical in avoiding
agency favoritism of a single interest group. Further, there is an
indication that industrial sensitive or highly regulated industries
are more likely to embrace voluntary programs. Probably, corporate
leaders should take it as a personal initiative to implement indus-
trial specific sustainability initiatives that can lead to better envi-
ronmental performance.
In conclusion, this study examines whether the adoption
voluntary environmental management systems enhance corporate
environmental performance among 852 firms in Kenya. The survey
results are then used in empirical methods to understand how both
ISO 14001 and UNGC relate to waste management activities of a
firm. Our study shows that indeed adoption of ISO 14001 and UNGC
lead to better environmental performance. This is partly by
encouraging member organizations to set environmental goals that
is helpful in focusing an organization’s internal activities towards
environmental management activities in presence of self-oversight.
Partly, also, due to members program training offered by the third
parties on aspects such as resource management that may translate
into better achieving of the program goals. Lastly, the observed
better environmental performance may be a case of increased in-
formation sharing of best practices or rigorous collaboration among
the participating firms. Of concern were the mixed findings with
ISO 14001. Our take is that the failure of third-party auditors, of ISO
14001 participants, to do public disclosures and a lack of solid
punitive framework may be facilitating general other than specific
improvements. Then, it is important to exercise greater monitoring
and enhance ambitious yet realistic goals. Interestingly, though
cautiously, we were able to show that intangible influences of
rewarding senior management encourages the adoption voluntary
programs. Finally, this research demonstrates the need for further
research to expand on the understanding of voluntary environ-
mental programs with suggested study implications.
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.
CRediT authorship contribution statement
Edward M. Mungai: Writing - original draft, Visualization,
Validation, Software, Resources, Project administration, Method-
ology, Investigation, Formal analysis, Data curation, Conceptuali-
zation. S. Wagura Ndiritu: Supervision, Writing - review & editing.
Tazeeb Rajwani: Supervision, Writing - review & editing.
Acknowledgments
We are humbled by the encouragement, motivation, and inputs
from peers at Strathmore University Business School.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jclepro.2020.121636.
References
Abbott, K.W., Snidal, D., 2010. International regulation without international gov-
ernment: improving IO performance through orchestration. In: The Review of
International Organizations, 5, pp. 315e344, 3.
Alauddin, M., Tisdell, C.A., 2006. Students’ Evaluation of Teaching Effectiveness:
what Surveys Tell and what They Do Not Tell. Economic Theory, Applications
and Issues. Working Paper 42.
Altonji, J., Elder, T., Taber, C., 2005. Selection on observed and unobserved variables:
assessing the effectiveness of catholic schools. J. Polit. Econ. 113 (1), 151e184.
https://doi.org/10.1086/426036.
Annandale, D., Morrison-Saunders, A., Bouma, G., 2004. The impact of voluntary
environmental protection instruments on company environmental perfor-
mance. Bus. Strat. Environ. 13 (1), 1e12.
Arevalo, J.A., Aravind, D., Ayuso, S., Roca, M., 2012. The Global Compact: an analysis
12 E.M. Mungai et al. / Journal of Cleaner Production 265 (2020) 121636
of the motivations of adoption in the Spanish context. Bus. Ethics Eur. Rev. 22
(1), 1e15.
Arimura, T.H., Darnall, N., Ganguli, R., Katayama, H., 2016. The effect of ISO 14001 on
environmental performance: resolving equivocal findings. J. Environ. Manag.
166, 556e566.
Arimura, T.H., Hibiki, A., Katayama, H., 2008. Is a voluntary approach an effective
environmental policy instrument? J. Environ. Econ. Manag. 55 (3), 281e295.
Arora, Seema, Cason, Timothy N., 1996. Why Do Firms Volunteer to Exceed Envi-
ronmental Regulations? Understanding Participation in EPA’s 33/50 Program.
Land Economics Vol. 72 (4). https://doi.org/10.2307/3146906.
Babakri, K.A., Bennett, R.A., Franchetti, M., 2003. Critical factors for implementing
ISO 14001 standard in United States industrial companies. J. Clean. Prod. 11 (7),
749e752.
Blackman, A., 2010. Alternative pollution control policies in developing countries.
Rev. Environ. Econ. Pol. 4 (2), 234e253.
Christmann, P., Taylor, G., 2006. Firm self-regulation through international certifi-
able standards: determinants of symbolic versus substantive implementation.
J. Int. Bus. Stud. 37 (6), 863e878.
Dingwerth, K., 2008. Private transnational governance and the developing world: a
comparative perspective. Int. Stud. Q. 52 (3), 607e634.
Eng Ann, G., Zailani, S., Abd Wahid, N., 2006. A study on the impact of environ-
mental management system (EMS) certification towards firms’ performance in
Malaysia. Manag. Environ. Qual. Int. J. 17 (1), 73e93.
Fei-Baffoe, B., Botwe-Koomson, G.K., Mensa-Bonsu, I.F., Appiah Agyapong, E., 2013.
Impact of ISO 14001 environmental management system on key environmental
performance indicators of selected gold mining companies in Ghana. J. Waste
Manag. 1e6, 2013.
Flankova, S., Tashman, P., Van Essen, M., Marano, V., 2018. A meta-analysis of the
effectiveness of voluntary environmental programs. Acad. Manag. Proc. 2018
(1), 14943.
Franchetti, M., 2011. ISO 14001 and solid waste generation rates in US
manufacturing organizations: an analysis of relationship. J. Clean. Prod. 19
(9e10), 1104e1109.
Gomez, A., Rodriguez, M.A., 2011. The effect of ISO 14001 certification on toxic
emissions: an analysis of industrial facilities in the north of Spain. J. Clean. Prod.
19 (9e10), 1091e1095.
Government of Kenya, 2018. National Climate Change Action Plan 2018-2022 (1).
Ministry of Environment and Forestry.
Halis, M., Halis, M., 2016. Relationship between environmental management sys-
tems implementation and environmental performance: findings from Turkish
EMS certificated businesses. Int. J. Org. Leadership 5 (20), 137e150.
Henri, J., Journeault, M., 2008. Environmental performance indicators: an empirical
study of Canadian manufacturing firms. J. Environ. Manag. 87 (1), 165e176.
Henriques, I., Sadorsky, P., 2007. Environmental management systems and prac-
tices: an international perspective. Environ. Pol. Corp. Behav.
Hertin, J., Berkhout, F., Wagner, M., Tyteca, D., 2008. Are EMS environmentally
effective? The link between environmental management systems and envi-
ronmental performance in European companies. J. Environ. Plann. Manag. 51
(2), 259e283.
Iwata, K., Arimura, T., Hibiki, S., 2010. An empirical analysis of determinants of ISO
14001 adoption and its influence on toluene emission reduction. JCER Econ J 62,
16e38.
Jaaffar, A.H., Amran, A., Rajadurai, J., 2018. The impact of institutional pressures of
climate change concerns on corporate environmental reporting practices: a
descriptive study of Malaysia’s environmentally sensitive public listed com-
panies. SAGE Open 8 (2), 215824401877483.
Khanna, M., 2002. Non-mandatory approaches to environmental protection. J. Econ.
Surv. 15 (3), 291e324.
Kleme
s, J.J., Varbanov, P.S., Huisingh, D., 2012. Recent cleaner production advances
in process monitoring and optimisation. J. Clean. Prod. 34, 1e8.
Kohna, P.T.B., 2004. Implementing global standards e the emerging role of the non-
state actor. Environmental Policy and Law 34, 260e265.
Kube, R., Von Graevenitz, K., Lo €schel, A., Massier, P., 2019. Do voluntary environ-
mental programs reduce emissions? EMAS in the German manufacturing
sector. SSRN Electron. J. Discussion Paper 107.
Lee, E., Jung, C.S., Hwang, M.S., 2015. Investigating supportive conditions for
participation in voluntary environmental programs. J. Environ. Plann. Manag.
59 (7), 1323e1340.
Liu, N., Tang, S., Zhan, X., Lo, C.W., 2018. Policy uncertainty and corporate perfor-
mance in government-sponsored voluntary environmental programs.
J. Environ. Manag. 219, 350e360. https://doi.org/10.1016/j.jenvman.2018.04.110.
Mathers, N., Fox, N., Hunn, A., 2007. Surveys and questionnaires: the NIHR RDS for
the east midlands/yorkshire & the humber. Retrieved from. https://www.rds-
yh.nihr.ac.uk/wp-content/uploads/2013/05/12_Surveys_and_Questionnaires_
Revision_2009.pdf/uploads/2013/05/12_Surveys_and_Questionnaires_
Revision_2009.pdf.
Nawrocka, D., Parker, T., 2009. Finding the connection: environmental management
systems and environmental performance. J. Clean. Prod. 17 (6), 601e607.
Nemati, M., Zheng, Y., Hu, W., 2016. ISO-14001 standard and firms’ environmental
performance: evidence from the U.S. Transportation equipment manufacturers.
SSRN Electron. J. 44 (2), 422e438.
Nguyen, Q.A., Hens, L., 2015. Environmental performance of the cement industry in
Vietnam: the influence of ISO 14001 certification. J. Clean. Prod. 96, 362e378.
Nishitani, K., Kaneko, S., Fujii, H., Komatsu, S., 2012. Are firms’ voluntary environ-
mental management activities beneficial for the environment and business? An
empirical study focusing on Japanese manufacturing firms. J. Environ. Manag.
105, 121e130.

Ortas, E., Alvarez, I., Garayar, A., 2015. The environmental, social, governance, and
financial performance effects on companies that adopt the united Nations
global Compact. Sustainability 7 (2), 1932e1956.
Qi, G., Zeng, S., Shi, J.J., Meng, X., Lin, H., Yang, Q., 2014. Revisiting the relationship
between environmental and financial performance in Chinese industry.
J. Environ. Manag. 145, 349e356.
Rasche, A., 2011. ‘A necessary supplement’ d what the united Nations global
Compact is (and is not). Corp. Soc. Human Rights Responsibil. 52e76. https://
doi.org/10.1057/9780230294615_3.
Rasche, A, de Bakker, F.G.A., Moon, J, 2013. Complete and Partial Organizing for
Corporate Social Responsibility. J. Bus. Ethics 115, 651e663. https://doi.org/
10.1007/s10551-013-1824-x.
Rivera, J, 2002. Assessing a voluntary environmental initiative in the developing
world: The Costa Rican Certification for Sustainable Tourism. Policy Sciences 35
(4), 333e360.
Russell, C. S., & Vaughan, W. J. (n.d.). The choice of pollution control policy in-
struments in developing countries: arguments, evidence and suggestions. The
International Yearbook of Environmental and Resource Economics 2003/2004.
doi:10.4337/9781781950173.00016.
Runhaar, H, Lafferty, H, 2009. Governing Corporate Social Responsibility: An
Assessment of the Contribution of the UN Global Compact to CSR Strategies in
the Telecommunications Industry. J. Bus. Ethics 84, 479e495. https://doi.org/
10.1007/s10551-008-9720-5.
Seiffert, M.E., 2008. Environmental impact evaluation using a cooperative model for
implementing EMS (ISO 14001) in small and medium-sized enterprises.
J. Clean. Prod. 16 (14), 1447e1461. https://doi.org/10.1016/j.jclepro.2007.10.001.
Shetty, S., Kumar, S., 2016. Are voluntary environment programs effective in
improving the environmental performance: evidence from polluting Indian
Industries. Environ. Econ. Pol. Stud. 19 (4), 659e676. https://doi.org/10.1007/
s10018-016-0168-z.
Tatoglu, E., Frynas, J.G., Bayraktar, E., Demirbag, M., Sahadev, S., Doh, J., Koh, S.C.,
2019. Why do emerging market firms engage in voluntary environmental
management practices? A strategic choice perspective. Br. J. Manag. 31 (1),
80e100. https://doi.org/10.1111/1467-8551.12351.
Testa, F., Rizzi, F., Daddi, T., Gusmerotti, N.M., Frey, M., Iraldo, F., 2014. EMAS and ISO
14001: the differences in effectively improving environmental performance.
J. Clean. Prod. 68, 165e173. https://doi.org/10.1016/j.jclepro.2013.12.061.
Tung, A., Baird, K., Schoch, H., 2014. The association between the adoption of an
environmental management system with organisational environmental per-
formance. Australas. J. Environ. Manag. 21 (3), 281e296. https://doi.org/
10.1080/14486563.2014.932721.
UNEP, 2010. Annual Report. UNEP.
Welch, Eric W., Hibiki, Akira, 2003. An institutional framework for analysis of
voluntary policy: the case of voluntary environmental agreements in Kita
Kyushu, Japan. J. Environ. Plann. Manag. 46 (4), 523e543. https://doi.org/
10.1080/0964056032000133143.
Yong, J.Y., Kleme
s, J.J., Varbanov, P.S., Huisingh, D., 2016. Cleaner energy for cleaner
production: modelling, simulation, optimisation and waste management.
J. Clean. Prod. 111, 1e16. https://doi.org/10.1016/j.jclepro.2015.10.062.
Zhang, B., Yang, S., Bi, J., 2013. Enterprises’ willingness to adopt/develop cleaner
production technologies: an empirical study in Changshu, China. J. Clean. Prod.
40, 62e70. https://doi.org/10.1016/j.jclepro.2010.12.009.
Zhou, R., Bi, X., Segerson, K., 2020. Evaluating voluntary environmental programs
with spillover effects. J. Assoc. Environ. Resour. Econ. 7 (1), 145e180. https://
doi.org/10.1086/705828.
Zobel, T., 2015. The impact of ISO 14001 on corporate environmental performance:
a study of Swedish manufacturing firms. J. Environ. Plann. Manag. 59 (4),
587e606. https://doi.org/10.1080/09640568.2015.1031882.
Zucker, L., 1987. Institutional theories of organization. Annu. Rev. Sociol. 13 (1),
443e464. https://doi.org/10.1146/annurev.soc.13.1.443.