Energy 68 (2014) 922e929

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Green businesses in a clean energy economy: Analyzing drivers

of green business growth in U.S. states
Hongtao Yi*
John Glenn School of Public Affairs, The Ohio State University, 1810 College Rd, Columbus, OH 43210, USA

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 30 November 2013
Received in revised form
9 February 2014
Accepted 13 February 2014
Available online 12 March 2014

In a clean energy economy, green businesses play a central role by utilizing renewable energy technologies and employing green labor forces to provide clean energy services and goods. This paper aims at
analyzing factors driving the growth and survival of green businesses in the U.S. states, with hypotheses
proposed on the impacts from clean energy policies and tax incentives, labor market conditions, and
economic and political environments. A xed effect regression analysis is applied with a panel data set of
48 continental states from 1998 to 2007 in the United States. The statistical analysis with a longitudinal
data set reveals that the adoption of renewable energy policies, the permission of renewable energy
credits imports, the stringency of minimum wage legislations, and presence of clean energy business
associations are the major driving forces of the green business development in the U.S. states.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Green economy
Green businesses
Green jobs
Low carbon society

1. Introduction
The clean energy economy is an explosively growing economic
sector over the last decade across the globe, as a response to global
climate change, depleting traditional energy sources, and the need
for industrial upgrade and structural transitions. For most countries, the most formidable challenge in mitigating climate change is
to overcome the nancial cost, especially when the country is
suffering economic recession [47]. Thus, it is essential to ensure the
robustness of a green economy so that the uses of renewable and
sustainable technologies bring about stable economic growth, two
indicators of which are job creation and business growth. In a clean
energy economy, green businesses play a central role by utilizing
renewable energy technologies and employing green labor forces
to provide clean energy services and goods. Utilizing different kinds
of renewable technologies, including biomass, solar thermal, energy efciency, heat pumps and distributed generation [5,12,19, 50],
green businesses are at the center of the clean energy economy.
Therefore, it is essential to understand the policy, economic and
political environments faced by the green businesses, which are the
backbones of a green economy. More importantly, it is important to
understand the driving factors underlying the growth and decline
of green businesses so that policy makers and stakeholders can
make timely adjustment to their policy tools.

* Tel.: 1 614 247 8791; fax: 1 614 292 2548.

E-mail addresses: yi.201@osu.edu, hongtaoyi08@gmail.com.
http://dx.doi.org/10.1016/j.energy.2014.02.044
0360-5442/ 2014 Elsevier Ltd. All rights reserved.

Over the past decade, the states have experienced a rapid

growth of renewable energy, with an accelerating rate over the last
few years. Net generation of electricity from renewable energy in
the U.S. grew from 79 million MWh (megawatthours) in 2003 to
218 million MWh in 2012 [16], compared with relatively slow
growth from 64 million MWh to 79 million MWh from 1990 to
2003. Renewable energy capacity in electricity generation varies
signicantly across the states, with many states having no renewable energy capacity and Texas and California having 7536 MW and
6168 MW in 2008 respectively.
Economic growth is closely tied to energy development and
consumptions [27,32]. The transition to a low-carbon, sustainable
and clean energy economy requires a systematic transition in the
whole economy, covering a whole range of economic sectors,
including manufacturing, utilities, services and training, transportation, agricultural and natural resources conservation
[15,29,45]. Therefore, the clean energy economy is not a single industry, but an overarching umbrella that covers burgeoning business activities that contribute to the conservation of energy and the
growth of alternative energy resources [26]. In the context of
United States, it has attracted much attention from academic
community, think tanks and the government, who have recently
come to a shared understanding of what constitutes a clean energy
economy. The Pew Charitable Trust denes the clean energy
economy as an economy that generates jobs, businesses and investments while expanding clean energy production, increasing
energy efciency, reducing greenhouse gas emissions, waste and
pollution, and conserving water and other natural resources [51],

H. Yi / Energy 68 (2014) 922e929

p. 5. Brookings Institution proposes that clean economy can be

measured by business establishment and jobs associated with
them. The clean economy produces goods and services with an
environmental benet or adds value to such products using skills or
technologies that are uniquely applied to those products [34].
The U.S. BLS (Bureau of Labor Statistics) denes green goods and
services as jobs in businesses that produce goods and provide
services that benet the environment or conserve natural resources [53].
The two pillars underlying the clean energy economy are business establishments and jobs. The creation of green jobs and the
growth of green businesses are both essential for clean energy
development. In the political discussions, much emphasis was
placed on the creation of green jobs, partly due to the fact that the
economy needs more jobs in an economic downturn. Academically,
many studies were conducted to evaluate the potential of job creation in the clean energy sectors. Ref. [46] estimated the job creation potential of solar PV installations [23]. reported that a 20%
national RPS (Renewable Portfolio Standard) by 2020 (85% biomass,
14% wind energy, 1% solar PV) would create a total employment of
163,669 for the United States. Ref. [54] forecasted that aggressive
energy efciency policies combined with a 30% national renewable
portfolio standards target in 2030 could create over 4 million jobs
by 2030 in the United States. Ref. [42] estimated that 400,000 new
green jobs would be created under full compliance with the
greenhouse gas emission regulations in California. Similar green
economy and green jobs studies were conducted in other national
contexts [3,7,9,25,28,30,33,38,44,52]. These studies are mostly exante evaluations of job creation potentials of relevant clean energy
technology and policy scenarios, using inputeoutput analyses.
Several think tanks make efforts to classify, inventory and
aggregate data on the number of green jobs that have actually
been created and maintained over the years in the United States
[18,34,51]. These reports provide detailed data and analysis on green
jobs at state and metropolitan levels, but no statistical analysis is
conducted to explain or analyze the factors affecting green job creations. Ref. [57] investigates policy and labor market factors underlying the distribution of green jobs across the U.S. metropolitan areas
and nds that metropolitan areas with clean energy policies, such as
RPS and other tax incentives are likely to have more green jobs.
Compared with the burgeoning literature on green jobs forecasting and evaluations, the evaluation of the growth trajectories of
green businesses is somehow absent from the literature, except for
a study by Ref. [6]. Given that both green jobs and green businesses
are necessary components of the clean economy, the growth and
survival of green businesses deserve academic attention and investigations. The growth of clean energy economy is not only about
creating green jobs, but also about the long-term survival of clean
energy businesses. Several differences between green businesses
and green jobs are worth discussion. First, green businesses and
green jobs may undergo different trajectories in the same area.
Although green businesses provide employment opportunities,
they may also make strategic responses to the policy, labor market,
economic and political environments by laying-off employees. Thus
a state or region may experience a growth in the number of green
businesses without the growth of green jobs, or vice versa. Therefore, studying the survival of green businesses is a different question from evaluating the green job creation.
Second, green businesses and green jobs serve slightly different
functions in the clean energy economy. Green jobs are essential for
contributing to the overall employment of the economy, since a low
unemployment rate is always preferred by policy makers. Green
businesses serve multiple purposes by providing tax revenues and
green services, in addition to the provision of employment opportunities. Policy makers are not only concerned with job creations,

923

but also overall scal health of the government, partially affected by

the tax revenues. Third, green businesses are more robust to
external policy, labor market, economic and political shocks
compared with the green employment. A green business can lay-off
workers when in business hardships, so that they can preserve itself and seize new opportunities for growth. Thus, to some extent,
the health and competitiveness of green businesses are essential
for ensuring the long-term green employment growth.
Given the unique roles of green businesses in the clean energy
economy, and that there is no previous study investigating the
factors underlying the growth and survival of green businesses in
the context of United States, this paper aims at analyzing the policy,
labor market, economic and political factors inuencing the survival of green businesses in the 48 continental states. The goal is to
theorize and test what factors are most critical for shaping the
patterns of green businesses so that policy makers and relevant
stakeholders are aware of the factors that can be manipulated to
adjust the overall trajectory in the green economy. The next section
features a theoretical framework developed to explain green
business survival. A xed effects model is then presented as the
modeling strategy, with descriptions of variables and data. The
results of the model are then discussed. The paper concludes with
policy implications.
2. A theoretical framework for green business survival
To explain the survival of green businesses in the context of state
economies, a four component framework is developed here as
presented in Fig. 1. In this framework, factors that could potentially
shape the business environment in the clean energy economy are
categorized into four major components: clean energy policies,
labor market conditions, economic environment and political
environment. The rst set of factors includes the adoption and
implementation of RPS (renewable portfolio standards) and tax
incentives, which stimulate the market demand for renewable
energy businesses. With policy-induced market demand for renewables and energy efciency, as well as other low-carbon services and goods, the green businesses take the opportunities to
meet these demands. The uctuations in clean energy policies
could result in market anxieties so that businesses reduce or delay
investment. In the most extreme cases, green businesses could go
bankrupt when policy support for the industry is reversed.
The second set of factors covers labor market conditions. For
green businesses, labor market conditions are essential for business
operations, as the cost of conducting business and the quality and
quantity of labor supplies uctuates with the rules and conditions
in the labor market. Three factors in the labor market are being
considered here: the stringency of minimum wage laws, unionization activities and average educational attainment. The rst two
factors are regulating the degree of freedom in hiring decisions and
average educational attainment is a measure of labor quality in the
region. The next important factor is economic environment, which
is critical for the growth of green businesses. Clean energy economy
is part of and shaped by the larger economic environment. The key
indicators of economic environment, GSP (Gross State Product) and
unemployment rate, are included here. The argument here is that
clean energy economy is embedded within a larger state economy
that provides market demand and relevant services for green
businesses.
Political factors are the last explanations of green business
growth. Many green businesses are start-ups with urgent need for
policy and investment support, and are potentially fragile when the
political support for the new businesses fades away. As a result,
green businesses are more vulnerable to the inuences from the
larger political environment. The social and political acceptability

924

H. Yi / Energy 68 (2014) 922e929

Fig. 1. A framework for green business survival.

of clean energy goods, service and products could directly affect the
market share of green businesses compared with other sectors of
the economy. The activities of interest groups are also essential in
that they can translate diffused demand for policies into concentrated business interests, and thus exert larger inuence in the
policy making process. In summary, this four-component theoretical framework provides an organizing structure for this paper. In
the next section, modeling strategies, together with variables and
data, are discussed in details.
3. Methods and data
3.1. Model
The purpose of this paper is to analyze policy, economic, political and labor market factors contributing to the growth and survival of green businesses in the 48 continental states from 1998
to 2007. A xed-effect regression model is employed to account
for unobserved state-level and time variations in green business
development. In mathematical formula, the model could be
formulated as:

Y s;t as st bX s;t s;t

where s is the state and t is the year of the observation. Ys,t is the
dependent variable, which is the logged total number of green
businesses in state s for year t. Xs,t are explanatory variables,
including variables on RPS and other tax incentives, labor market,
business competition, economic environment, tax burden and
other costs, and political environment. as is a state-specic intercept. st denotes a vector of time dummies that control for time xed
effect. s,t varies independently across time and states.
The use of xed effect model presents several advantages. First,
it generates consistent estimation even when explanatory variables
and the error term are correlated. This means the validity of the
estimates will be not affected by omitted variables. Second, it assumes that states are heterogeneous and that the state-specic
intercept determines the baseline for green businesses in that
state. This is consistent with the empirical observation that states

are different in their overall level of green businesses to start with.

For example, the number of green businesses in Indiana was 1135 in
1998, compared with 2111 in Illinois.
3.2. Dependent variable: number of green businesses
The dependent variable of this study is the number of green
businesses in the states from 1998 to 2007. Overtime, the number
of business establishment captures the combined effects of business births and deaths. This measure was coded from a report by
the Pew Charitable Trust published in Ref. [51], titled The Clean
Energy Economy: Repowering Jobs, Businesses and Investments
across America, which denes clean energy businesses as establishments that expand clean energy production, increase energy
efciency, reduce greenhouse gas emissions, waste and pollution,
and conserve water and other natural resources [51], p. 11. The
number of green businesses is calculated by adding the number of
establishments that fall into 74 categories of SIC (Standard Industrial Classication) codes associated with these categories, based
on the National Establishment Time Series database by Dun &
Bradstreet.
The following are a few examples of green businesses as documented by the report. Gamesa, a clean energy rm producing wind
turbine was established in Pennsylvania in 2005, employing around
1000 workers as of 2009. The key factors affecting the location
choice of this rm were the states bipartisan support for renewable
energy, a skilled workforce, and aggressive renewable portfolio
standards [51]. An example of energy efciency rm is Johnson
Controls, which provides building-related energy efciency services
and employs more than 40,000 employees for efciency services
[51].
The number of green businesses increased from 61,689 to
68,203 during the decade of 1998e2007, a 9.1 percent growth. At
the state level, there are wide variations among the individual state
in the green business development trajectories. Some states, like
Wisconsin and Pennsylvania, experienced relatively stable green
economic growth. Wisconsin witnessed a modest increase in the
number of green businesses from 1249 to 1294. Other states, like
California (14.6% growth rate) and Colorado (21.5% growth rate),

H. Yi / Energy 68 (2014) 922e929

have undergone signicant boom in green business growth. However, in a few states, for example Oklahoma, West Virginia and New
Jersey, the total number of green businesses experienced moderate
declines, indicating that there are more green rms that went
bankrupt than survived. To explain the variation among the states
in green business patterns is the main purpose of this study. The
following sections focus on the factors that might explain the
diverging patterns of state green business growth.
3.3. RPS and tax incentives
RPS is one of the major policy instruments states have adopted
to stimulate the renewable energy industry (Table 1). Among the
various policy instruments, RPS is argued to be the most effective
for promoting renewables [36]. Rader and Norgaard also argue that
RPS is the most efcient means of correcting market imperfections
and for moving toward sustainability [41], p.38. RPS is not a singlepurpose policy instrument that only addresses climate change and
renewable energy, but it is a multiple-purpose instrument that
encompasses additional goals to stimulate economic development.
The rationales for the adoption of RPS are based on economic
development instead of climate protection in some states [39]. For
example, the RPS legislation in Texas does not emphasize its impact
on climate, but instead job creation and energy supply diversity.
In the literature, it is argued that RPS can also produce substantial economic benets to states. Ref. [40] argues that RPS is
economically benecial to the adopting state and consistent with
the goals of economic development. Ref. [55] reports that RPS
motivates renewable energy development. Ref. [58] argue that RPS
is a substantive rather than symbolic policy in that RPS signicantly and positively contributes to the development of renewable
energy. The logical link between renewable energy development
and economic development is based on the jobs created by the
renewable energy industry. Although some study has shown that
the adoption of RPS may not entirely be driven by a motivation to
create jobs [56], but other studies demonstrated that renewable
energy development has a great potential in creating green jobs
[54].
It can be hypothesized that RPS could play a signicantly positive role in inducing renewable technology adoptions in the states.
In addition, states differ in their RPS rules and these rules may bring
about spill-over effects of green economic development. Some
states allow their utilities to buy RECs (Renewable Energy Credits)
from other states to fulll in-state RPS goals, while other states do
not. States are also varying in the rules regulating whether the
electric utilities are allowed to export RECs to other states. Allowing
utilities to purchase RECs from outside of the state would

Table 1
Timeline of state adoption of RPS.
Year of adoption

States

1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008

Minnesota

Massachusetts, Nevada
Connecticut
Maine, New Jersey, Texas, Wisconsin
Hiwaii
California
Colorado, Maryland, New Mexico, New York, Pennsylvania
Delaware, Montana, Vermont
Arizona, North Dakota, South Dakota, Washington
Illinois, New Hampshire, North Carolina, Oregan, Virginia
Michigan, Missouri, Utah

925

disincentivize in-state renewable energy generation, and thus

affect the growth and survival of green businesses in the state. But
from the perspective of the region or nation, green businesses
would benet from green energy policies adopted in other states. If
there is a race to the top phenomenon, as demonstrated in the
adoption of clean energy policies in the states, then green businesses across the country would be the beneciaries. On the other
hand, forbidding the state renewable energy generators to export
RECs theoretically would help keep the economic development
benets within the state.
To measure RPS, a binary variable is used to indicate whether a
state adopts mandatory RPS in a given year, with (1) indicating a
state adopts mandatory RPS in that year, and (0) indicating a state
adopted voluntary or no RPS. Two dummy variables were created to
measure whether a state allows utility companies to purchase/
export RECs from/to other states separately. All data on RPS designs
were collected from DSIRE (Database for State Incentives for Renewables and Efciency).
Another important aspect of the state renewable energy policies
is the rise of nancial incentives, most notably tax incentives,
including property tax, sales tax, corporate income tax and personal income tax incentives. In theory, tax incentives help adjust
market imperfections by stimulating the development of renewables. Many states adopt multiple policy incentives for renewables
[10]. For clean energy businesses, the impacts of tax incentives are
signicant. Tax incentives can enlarge the market of clean energy
and energy efciency for the clean energy businesses. For example,
a property tax incentive for the solar installation would incentivize
individuals and companies to adopt solar PV panels that they
otherwise would not. Obviously, businesses that specialize in solar
PV panel manufacturing and installations are going to benet from
the increased market penetration of renewable technology, incentivized by the use of tax incentives.
Tax incentives are measured by an index of adoptions of
corporate income tax incentives, property tax incentives, personal
income tax incentives and sales tax incentives. For each state, the
tax incentive index score in a given year is equal to the currently
effective tax incentives in a state. Therefore, this variable is also
time-varying. The data on tax incentives were also collected from
DSIRE.
3.4. Labor market
The development and survival of green businesses are closely
tied to labor market supplies and government policies regulating
labor market. The factors of interest are minimum wage legislations, unionization of labors and average education attainment of
the labors. These factors affect the cost and availability of workforce
in the labor market.
In the United States, in addition to federal minimum wage
laws, states can also pass legislations to raise the minimum wages
for their individual state. It is generally argued that stronger
minimum wage law may lead to increased unemployment rates,
as it restricts the ability of both employees and employers to
negotiate contracts [8,24,48]. The minimum wage laws negatively
affect people who are new to the labor market and relatively
unskilled. Green businesses need a wide variety of occupations to
perform business functions, with some occupations better paid
than others. According the Bureau of Labor Statistics of the U.S
[53]., a green job in a green business may be a job in transportation and material moving, installation, maintenance, and
repair, construction and extraction, along with employees in the
occupations of management and nancial operations. Some occupations are not paid as well. For example, the transportation
and material moving occupation accounts for 27% of employment

926

H. Yi / Energy 68 (2014) 922e929

in the green businesses in 2011 [53] with an average salary of

$36,720, a wage below the national occupational average of
$45,230 of that year. This means that green businesses are not
exempt from the inuence from minimum wage laws, and that the
variation among states in the stringency of the minimum wage
laws may affect the protability and thus the survival of green
businesses. This variable is measured by the annual income
earned by someone working at the minimum wage as a ratio of
per capita GSP in the state, collected from Economic Freedom of
North America [2].
The second factor in the labor market is the unionization of the
labor. In labor economics, it is commonly understood that unionization activities impose costs to employers by raising the wages
of the unionized members, limiting the hiring and ring decisions
of the businesses, and changing the structure of wage payments
across occupations. These activities could lead to reduced employment, decline in organizational performance and shutdown of
businesses [1,17,21,43]. However, the impact of unions has seen a
decline in the last decades, and some studies have shown little
economic impacts of unionization on wages, unemployment and
business survivals, because unions are not successful in securing
wage gains [14]. A variable of unionization activities is included to
test and control for the inuence of unions on the survival of green
businesses. This variable is measured as a percentage of unionized
workers in a state, collected from Economic Freedom of North
America [2].
The third labor market characteristic is the average education
levels of the labor. Clean energy industries need different levels of
specialized and well trained professionals to complete various tasks
involved in energy auditing, solar panel installation, thin lm
production, clean energy consulting and education etc. Although it
is not clear whether jobs in clean energy sectors require more skills
than existing traditional energy sectors [51], a study shows that
average education attainment is positively associated with green
jobs at the metropolitan level in the United States [57]. The lack of
qualied labor could restrain the development of any business,
including a newly developed green business. Therefore, it is very
crucial that green business locate in a state with enough qualied
workforce, which is in urgent demand in a green economy [35]. As a
result, the average education level in the states needs to be taken
into account for local labor market conditions in the regressions
models.
3.5. Economic environment
Economic environment shapes the growth of clean energy industry and related services in the states. The rst factor is the
overall level of local economic productivities as measured by per
capita GSP (Gross State Product) in the state. The survival of green
businesses depends upon a sufcient level of local demand. The
clean energy R & D and manufacturing industries that have a national or global market also need to locate in an economically
vibrant metropolitan area to have access to other economic resources, such as nancial, legal and consulting services, to achieve
localization economies [22]. The higher the GSP level, the larger the
market size in the state, creating a sufciently large market to
achieve economies of scale for clean energy and related service
industries to develop and grow. GSP is measured in millions of
current dollars per person, collected from BEA (Bureau of Economic
Analysis).
The second factor related to the survival of green businesses is
the unemployment rate in the state. The unemployment rate
measures the local economic health, which could signicantly
affect the economic environment in which clean energy industries operate. As documented by Ref. [57], two potential

mechanisms could come into play. When the unemployment

rate is high, it could signal a stressed state economy within
which clean energy industry is nested. High unemployment rate
would lead to reduced disposable income for consumers and the
demand for clean energy products and services, and thus
threatens the survival of some green businesses. On the other
hand, clean energy industries are found to be counter cyclical to
the general economic patterns [51], so that they are seen as
employment solutions to high unemployment rates. The data on
unemployment rate is collected from U.S. BLS (Bureau of Labor
Statistics).
3.6. Political environment
The success and survival of green businesses also depend on the
political environment that they are embedded in. The political
environment not only determines whether they will get policy
support from the state legislature [11], but also inuences the social
discourse and acceptance over the role of green businesses in the
state economy. The rst factor of the political environment is the
general citizen ideology in the state. The citizen ideology index
measures the mean position of citizens on a liberal-conservative
continuum [4]. It is generally argued that liberal political ideology
is associated with green policies and renewable energy programs
[31,49,56]. The data on citizen ideology index were collected by
William Berry and Richard Fording. Updated measures were
extracted from Fordings University of Kentucky data archive
website.
The second factor is the activities of clean energy and energy
conservation business associations as interest groups. Interest
group mobilization and advocacy is very important for the survival
of green businesses, in that better organized interest groups are
potentially more successful in securing favorable policy treatment
from the regulators. This is especially important given that many
clean energy and energy efciency businesses rely on tax incentives, rebates and loans to reduce the cost of operations and
make their services and goods protable. The formation of trade
associations could enhance information sharing among green
businesses, and enhance the ability of clean energy industries to get
their voice heard by the policy makers and actively shape favorable
policy environments. The activities of clean energy interest groups
are measured by the number of clean energy and energy conservation non-prot organizations in a state for a given year, which is
coded from the database of NCCS (National Center for Charitable
Statistics).
Federal, state and local governments levy taxes to support
different functions of the government. The variety of taxes includes
corporate income tax, sales tax, property tax, personal income tax,
contributions to social security insurance etc. All kinds of taxes are
relevant for business survival, but indirect tax revenues such as
property taxes, employment insurance, workers compensation,
and various pension plans are especially important for the businesses, as they constitute the cost function of the business operations. It can be expected that when businesses have to pay more
indirect taxes, their cost of doing business increases and their
likelihood of survival decreases. States vary in the collection of
indirect tax revenues due to varying stringency of legislations
across the states. The indirect tax burden is measured as indirect
tax revenue as a percentage of GSP, collected from Economic
Freedom of North America [2].
A set of control variables, including the ratio of natural gas
generation capacity, median household income (logged) and electricity price (logged), are also included in the analysis. Total number
of business establishments (logged) is also included as a baseline of
businesses in the state.

H. Yi / Energy 68 (2014) 922e929

Table 2
Descriptive statistics.
Variable

Observation Mean Standard Min

deviation

Green businesses (logged)

Renewable portfolio standards
Renewable credits to other states
Renewable credits from other
states
Tax incentives
Average educational attainment
Minimum wage legislation
Union density
Per capita GSP
Unemployment rate
Total tax burden
Citizen ideology
Energy NGOs (logged)
Size of natural gas industry
Median income (logged)
Total business establishments
(logged)
Electricity price

480
480
480
480
480
480
480
480
480
480
480
480
480
480
480
480
480

6.78
0.29
0.02
0.03

Max

0.91
0.45
0.12
0.16

4.80
0
0
0

9.24
1
1
1

1.28 1.19
25.79 5.39
32.02 5.83
12.79 5.16
36.90 7.84
4.61 1.08
27.09 3.05
50.66 15.04
1.23 0.97
0.29 0.22
1.63 0.17
11.49 0.89

0
14.60
17.60
3.30
21.58
2.20
17.31
8.45
0
0
10.07
9.79

4
49.10
49.70
27.70
69.73
8.10
37.19
95.97
4.14
0.98
11.50
14.03

2.11

3.72

2.71

0.31

4. Results and discussions

The descriptive statistics for all the variables are presented in
Table 2. A series of unit root tests are performed to test the stationarity of the dependent variable. If the panel is stationary, then a
static panel data method should be applied, otherwise a dynamic
specication should be used. Two kinds of tests are performed,
Augmented DickeyeFuller approach [13] and PhillipsePerron
approach [37], with different specications on lags, a linear trend,
or a drift for all three dependent variables. All specications reject
the null hypothesis of unit-root. Therefore, the dependent variable
is panel stationary, and thus xed-effect or random effect models
are applicable.
A Hausman test [20] is conducted to test the specication of
xed-effect versus random-effect models. The null hypothesis
that the difference of the coefcients estimated by the two specications is not systematic is rejected, indicating the choice of a
xed-effect model (c2(17 d.f.) 287.77). To control for heteroskedasticity, the standard errors reported are clustered robust
standard errors. The inclusion of time dummies could control for
the inuence of federal clean energy policies, such as production
tax credits and other incentive programs. Time dummies could also
control for unobserved factors inuencing green business development over time, such as the evolution of technology and consumption patterns.

927

Table 3
Determinants of green businesses dependent variable: Number of green businesses
(logged).
Variables

Coefcient

Standard errors

Renewable portfolio standards

Renewable credits to other states
Renewable credits from other states
Tax incentives
Average educational attainment
Minimum wage legislation
Union density
Per capita GSP
Unemployment rate
Total tax burden
Citizen ideology
Energy NGOs (logged)
Size of natural gas industry
Median income (logged)
Total business establishments (logged)
Electricity price
Constant
Observation
State xed effect
Year xed effect
R-squared (within)
R-squared (between)
R-squared (overall)

0.020**
0.014
0.017**
0.004
0.0002
0.005***
0.004
0.003*
0.004
0.001
0.0006
0.045***
0.078
0.004
0.016
0.006
6.683***
480
Yes
Yes
0.542
0.517
0.407

0.009
0.021
0.007
0.004
0.001
0.001
0.003
0.002
0.004
0.003
0.0005
0.012
0.052
0.042
0.021
0.024
0.512
e
e
e

*p < 0.10; **p < 0.05; ***p < 0.01 (two-tailed).

Two additional variables are used to measure the REC requirements in the RPS, one for REC imports, and one for REC exports. The coefcient for the variable indicating whether a state
accepts REC imports from other states is negative and signicant.
Specically, the magnitude of the coefcient means that the
number of green businesses in a given state decreases by 1.7% on
average with the adoption of RPS allowing REC imports over time,
controlling for other factors. This conrms the hypothesis presented earlier that the economic impact of RPS may spill over to the
neighboring states, and that implementing an RPS allowing REC
imports are not really beneting instate business growth and survival. However, the coefcient for REC exports are not signicant,
not supporting the hypothesis that state green businesses would
benet from a RPS that allows REC exports.
The impact of tax incentives on the growth and survival of green
businesses was hypothesized to be positive. Surprisingly, the result
does not lend support to this hypothesis. To tentatively explain the
insignicance of the tax incentives, it can be argued these tax incentives may be conducive for business market expansion, but are
not necessary for the survival of the businesses.
4.2. Labor market: minimum wage legislation reduces likelihood of
business survival

4.1. Renewable energy policy: RPS design matters

The results of the analysis are mixed on the hypotheses discussed earlier, as presented in Table 3. The rst set of results to be
discussed is the clean energy policy variables. This study measures
the RPS policies with three variables, and each covers a unique
dimension. The coefcient of RPS, measuring whether there is a RPS
in the state, is positive and statistically signicant at 0.05 level. The
magnitude of the coefcient means that the number of green
businesses in a given state increases by 2% on average with the
adoption of RPS over time. This indicates that renewable portfolio
standard, as the most important state level clean energy legislation,
is driving the growth of green businesses in the states. As the
compliance ratio of RPS is very high in most states over the years
[35], the positive impacts of RPS on clean energy business growth is
within expectations.

Three variables were included to measure labor market conditions: minimum wage legislation, union density and average education attainment. The impact of minimum wage legislation on the
number of green businesses was hypothesized to be negative, with
more stringent minimum wage laws decreasing the number of
green businesses. The result shows that the coefcient for minimum wage law in the states is negative and statistically signicant
at 0.01 level. To interpret the coefcient, there is a 0.5 percent
decrease in the number of green businesses associated with 1
percent increase in the state minimum wage, controlling for other
factors. This result is consistent with what was found in labor
economics literature that minimum wage legislation tends to
reduce employment and potentially affect business growth.
The other two variables on the labor conditions are union density and average education attainment. The result of the analysis

928

H. Yi / Energy 68 (2014) 922e929

indicates no signicant effects. The insignicance of union density

conrms the recent observation from Ref. [14] that unions have lost
their bargaining power in recent decades in securing wage gains.
When unions are no longer capable of bargaining for higher salaries
for unionized workers, their inuence on business survival reduces
signicantly. The insignicant coefcient of average educational
attainment might indicate that average education in the state has a
larger inuence over the growth of green jobs than the growth of
green businesses.
4.3. Political environment: business associations help preserve
green businesses
Regarding the inuence of political environment, the role of
citizen ideology and advocacy activities in clean energy business
associations were discussed. The coefcient of citizen ideology was
not signicant, indicating that there is no difference between liberal and conservative states in the trajectory of green business
growth. The positive and statistically signicant coefcient of clean
energy NGOs indicates that the presence of strong advocacy coalitions can have a positive impact on green business growth. To put
the size of the coefcient into perspective, as the number of clean
energy business associations increases by 1 percent, the number of
green businesses increases by 4.5 percent. This means that state
clean energy businesses that are successful in taking collective efforts are more likely to experience growth compared with those
that rely on an individualistic approach.
The results for economic environment are either insignicant or
signicant in opposite directions, suggesting a counter cyclical
pattern of green economic development. The coefcient of per
capita GSP is negative and signicant, contrary to the hypothesis.
This means that states with higher per capita GSP are going to have
fewer green businesses. The coefcients of unemployment rate and
total indirect tax burden fail to achieve signicance. This indicates
that clean energy economy is somehow a unique sector following a
slightly different trajectory compared with the overall state economy. This is consistent with observation by Ref. [34] that the clean
energy economy generally outperforms the overall economy during
the recession. The size of natural gas industry, median household
income, electricity price and total business establishment in the
state were included as controls, but none of them were signicant
in the results.
5. Conclusions
Clean energy has become a fast growing economy in the United
States in the last two decades. The development of the clean energy
economy presents interesting policy questions, among which the
growth and survival of green businesses being the most fundamental ones. This paper aims at analyzing factors driving the growth
and survival of green businesses in the states, with hypotheses
proposed on clean energy policies and tax incentives, labor market
conditions, economic environment and political environment. The
statistical analysis with a longitudinal data set reveals that the
adoption of renewable energy policies, the permission of REC imports, the stringency of minimum wage legislations, and presence of
clean energy business associations are the major driving forces of
the green business development.
Multiple policy implications can be drawn from this analysis.
First, maintaining and strengthening current RPS legislations are
essential for keeping the current momentum of green business
development in the states. As shown in this analysis, a state with
RPS is more likely to have a strong and robust green economy, in
which green businesses and industries have relatively stable
growth trajectories over time. Given the recent legislative efforts in

multiple states to repeal RPS (for example Ohio, Texas, and Virginia), the most conducive policy environment for clean energy
may be subject to change or reversal.
Second, the economic externalities of the policy impacts evidenced in this study indicate the need for inter-state collaborations
in renewable energy policies. Ideally, the economic externality of
the policies, also known as policy leakage, can be xed by a national
renewable portfolio standard. Given that the federal government in
the United States failed to pass a national RPS in 2009, inter-state
policy coordination is needed to balance the inter-state differences in clean energy development.
For renewable energy businesses who would like to keep a long
term sustainable growth, their best strategy would be to work with
each other through coalition building. Sharing policy information
and coordinating collective actions among the newly born green
businesses are essential for their survival and long term mutuallybenecial development. The recent trend in the increasing formation of clean energy and energy conservation business associations
indicates that green businesses are moving in the right direction.

References
[1] Abowd JM. The effect of wage bargains on the stock market value of the rm.
Am Econ Rev 1989;79:774e800.
[2] Ashby NJ, Bueno A, McMahon F. Economic Freedom of North America 2011.
Vancouver, BC: Fraser Institute; 2011.
[3] Barbier EB, Markandya A. A new blueprint for a green economy. Routledge;
2013.
[4] Berry WD, Ringquist EJ, Fording RC, Hanson RL. Measuring citizen and government ideology in the American states, 1960-93. Am J Pol Sci 1998;42:327e
48.
[5] Blass V, Corbett CJ, Delmas MA, Muthulingam S. Top management and the
adoption of energy efciency practices: evidence from small and mediumsized manufacturing rms in the US. Energy 2013;65:560e71.
[6] Bowen WM, Park S, Elvery JA. Empirical estimates of the inuence of
renewable Energy portfolio standards on the Green economies of states. Econ
Dev Q 2013;27:338e51.
[7] Cai W, Wang C, Chen J, Wang S. Green economy and green jobs: myth or
reality? the case of Chinas power generation sector. Energy 2011;36(10):
5994e6003.
[8] Card D. Using regional variation in wages to measure the effects of the federal
minimum wage. No. w4058. National Bureau of Economic Research; 1992.
[9] Chakrabarty S, Islam T. Financial viability and eco-efciency of the solar home
systems (SHS) in Bangladesh. Energy 2011;36(8):4821e7.
[10] Ciocirlan CE. Analysing preferences towards economic incentives in
combating climate change: a comparative analysis of US states. Clim Policy
2008;8:548e68.
[11] Coley JS, Hess DJ. Green energy laws and Republican legislators in the United
States. Energy Policy 2012;48:576e83.
[12] Dehghanian P, Hosseini SH, Moeini-Aghtaie M, Arabali A. Optimal siting of DG
units in power systems from a probabilistic multi-objective optimization
perspective. Int J Electr Power Energy Syst 2013;51:14e26.
[13] Dickey David A, Fuller Wayne A. Likelihood ratio statistics for autoregressive
time-series with a unit root. Econometrica 1981;49:1057e72.
[14] DiNardo J, Lee DS. Economic impacts of new unionization on private sector
employers: 1984e2001. Q J Econ 2004;119:1383e441.
[15] da Graa Carvalho M, Bonifacio M, Dechamps P. Building a low carbon society.
Energy 2011;36(4):1842e7.
[16] Energy Information Administration. Electric power monthly http://www.eia.
gov/electricity/monthly/pdf/epm.pdf; 2014.
[17] Freeman RB, Medoff JL. What do unions do. Ind Labor Relat Rev 1985;38:244e
63.
[18] Global Insight. U.S. metro economies: current and potential green jobs in the
U.S. economy. In: The United States Conference of Mayors; 2008.
[19] Gonzlez-Limn JM, Pablo-Romero MDP, Snchez-Braza A. Understanding
local adoption of tax credits to promote solar-thermal energy: Spanish municipalities case. Energy 2013;62:277e84.
[20] Hausman Jerry A. Specication tests in econometrics. Econometrica 1978;46:
1251e71.
[21] Hirsch BT, Schumacher EJ. Unions, wages, and skills. J Hum Resour 1998;33:
201e19.
[22] Ihlanfeldt K. The importance of the central city to the regional and national
economy: a review of the arguments and empirical evidence. Cityscape J
Policy Dev Res 1995;1:125e50.
[23] Kammen DM, Kapadia K, Fripp M. Putting renewables to work: how many
jobs can the clean energy industry generate?. RAEL Report. University of
California Berkeley; 2006.

H. Yi / Energy 68 (2014) 922e929

[24] Katz LF, Krueger AB. The effect of the minimum wage on the fast food industry. No. w3997. National Bureau of Economic Research; 1992.
[25] Lambert RJ, Silva PP. The challenges of determining the employment effects of
renewable energy. Renew Sustain Energy Rev 2012;16:4667e774.
[26] Li H, Bao W, Xiu C, Zhang Y, Xu H. Energy conservation and circular economy
in Chinas process industries. Energy 2010;35(11):4273e81.
[27] Lin B, Liu X. Dilemma between economic development and energy conservation: Energy rebound effect in China. Energy 2012;45:867e73.
[28] Lin YE, Chen PY, Chen CC. Measuring green productivity of country: a generlized metafrontier Malmquist productivity index approach. Energy 2013;55:
340e53.
[29] Lior N. Sustainable energy development: the present (2009) situation and
possible paths to the future. Energy 2010;35(10):3976e94.
[30] Lund H, Hvelplund F. The economic crisis and sustainable development: the
design of job creation strategies by use of concrete institutional economics.
Energy 2012;43(1):192e200.
[31] Matisoff DC. The adoption of state climate change policies and renewable
portfolio standards: regional diffusion or internal determinants? Rev Policy
Res 2008;25:527e46.
[32] Moe E. Energy, industry and politics: energy, vested interests, and long-term
economic growth and development. Energy 2010;35(4):1730e40.
[33] Moreno B, Lpez AJ. The effect of renewable energy on employment. The case
of Asturias (Spain). Renew Sustain Energy Rev 2008;12:732e51.
[34] Muro M, Rothwell J, Saha D. Sizing the clean economy: a national and regional
green jobs assessment. Brookings Institution; 2011.
[35] National Research Council. Emerging workforce trends in the U.S. Energy and
mining industries: a call to action. Washington, DC: The National Academies
Press; 2013.
[36] Palmer K, Burtraw D. Cost-effectiveness of renewable electricity policies.
Energy Econ 2005;27:873e94.
[37] Phillips PC, Perron P. Testing for a unit root in time series regression. Biometrika 1988;75(2):335e46.
[38] Qi T, Zhou L, Zhang X, Ren X. Regional economic output and employment
impact of coal-to-liquids (CTL) industry in China: an inputeoutput analysis.
Energy 2012;46:259e63.
[39] Rabe BG. Statehouse and greenhouse: the emerging politics of American
climate change policy. Washington, DC: Brookings Institution; 2004.
[40] Rabe BG. Race to the top: the expanding role of U.S. state renewable portfolio
standards. Sustainable Dev. L. & Poly 2006;7:10e6.
[41] Rader NA, Norgaard RB. Efciency and sustainability in restructured electricity
markets: the renewables portfolio standard. Electr J 1996;9:37e49.

929

[42] Roland-Holst D. Energy efciency, innovation and job creation in California.

UC Berkeley: Department of Agricultural and Resource Economics; 2008.
http://escholarship.org/uc/item/7qz3b977 [accessed 01.12.11].
[43] Ruback RS, Zimmerman MB. Unionization and protability: evidence from the
capital market. J Polit Econ 1984;92:1134e57.
[44] Sastresa EL, Usn AA, Bribin IZ, Scarpellini S. Local impact of renewables on
employment: assessment methodology and case study. Renew Sustain Energy
Rev 2010;14:679e90.
[45] Scarpellini S, Valero A, Llera E, Aranda A. Multicriteria analysis for the
assessment of energy innovations in the transport sector. Energy 2013;57:
160e8.
[46] Singh V, Fehrs J. The work that goes into renewable energy. Washington, DC:
Renewable Energy Policy Project; 2011. http://www.repp.org/articles/static/1/
binaries/LABOR_FINAL_REV.pdf [accessed 01.12.11].
[47] Stern N. Review on the economics of climate change. London HM Treasury;
2006.
[48] Stigler GJ. The economics of minimum wage legislation. Am Econ Rev
1946;36:358e65.
[49] Stoutenborough JW, Beverlin M. Encouraging pollution-free energy: the
diffusion of state net metering policies. Soc Sci Q 2008;89:1230e51.
[50] Taibi E, Gielen D, Bazilian M. The potential for renewable energy in industrial
applications. Renew Sustain Energy Rev 2012;16(1):735e44.
[51] The Pew Charitable Trusts. The clean energy economy: repowering jobs,
businesses and investments across America. The Pew Charitable Trusts; 2009.
[52] Tourkolias C, Mirasgedis S. Quantication and monetization of employment
benets associated with renewable energy technologies in Greece. Renew
Sustain Energy Rev 2011;15:2876e86.
[53] U.S. Bureau of Labor Statistics. Occupational employment and wages in green
goods and services summary. U.S. Census Bureau; 2011. http://www.bls.gov/
news.release/ggsocc.nr0.htm [accessed 20.10.13].
[54] Wei M, Patadia S, Kammen DM. Putting renewables and energy efciency to
work: how many jobs can the clean energy industry generate in the U.S.
Energy Policy 2010;38:919e31.
[55] Wiser R. Renewable portfolio standards in the United Statesea status report
with data through 2007. Lawrence Berkeley National Laboratory; 2008.
[56] Yi H, Feiock RC. Policy tool interactions and the adoption of state renewable
portfolio standards. Rev Policy Res 2012;29:193e206.
[57] Yi H. Clean energy policies and green jobs: an evaluation of green jobs in U.S.
metropolitan areas. Energy Policy 2013;56:644e52.
[58] Yin H, Powers N. Do state renewable portfolio standards promote in-state
renewable generation? Energy Policy 2010;38:1140e9.