Journal of Economics and Business 90 (2017) 65–76

Contents lists available at ScienceDirect

Journal of Economics and Business

That is what friends do: employee friendliness and

innovation夽
Hari P. Adhikari a , Wonseok Choi b , Nilesh B. Sah c,∗
a
Assistant Professor, Department of Economics, Finance and Information Systems, College of Business, Embry-Riddle Aeronautical
University, 600 S. Clyde Morris Blvd., Daytona Beach, FL 32114, United States
b
Doctoral Student, Department of Economics and Finance, College of Business and Entrepreneurship, The University of Texas Rio Grande
Valley, 1201 West University Dr., Edinburg, TX 78539-2999, United States
c
Assistant Professor, Austin E. Cofrin School of Business, University of Wisconsin – Green Bay, 2420 Nicolet Drive, Green Bay, WI 54311,
United States

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

Article history: Recent literature suggests that labor related issues can impact corporate innovation. In
Received 10 November 2015 this study we hypothesize and find that firms with congenial work environments innovate
Received in revised form 22 October 2016 more and have greater innovative efficiency. Our results also suggest that cash profit shar-
Accepted 27 October 2016
ing and employee involvement have a positive bearing whereas union relationships have
Available online 31 October 2016
a negative impact on innovation output of a firm. In addition, subsample analysis reveals
that employee-friendly policies are more important in industries having higher employee
JEL classification:
power. Our results are consistent with the logic that happy and satisfied employees tend to
G30
be more productive than the unhappy and dissatisfied ones. Our work suggests a channel
J24
J28 through which satisfied employees contribute towards increasing firm value − by increas-
O32 ing the innovation productivity and efficiency of the firm. So, firms must make continued
efforts towards creating and maintaining a friendly work environment.
Keywords: © 2016 Elsevier Inc. All rights reserved.
Innovation
Corporate environment
Employee friendliness
Labor productivity

1. Introduction

“Imagine a world where most organizations were the best place to work. Imagine what we could be getting done on the planet
if it were true” − Karen May, VP of people development, Google.
Modern day corporations are trying to ensure that all employees feel a sense of purpose in their work and that they are
satisfied and happy, as happy employees tend to be more productive than the unhappy ones (Oswald, Proto, & Sgroi et al.,
2009). As such, employee satisfaction is a very broad term and it can be considered as a concoction of several stand-alone

夽 Much of this paper was completed when the authors were doctoral students at University of South Florida and University of North Texas. The authors
would like to thank the reviewers, discussants and conference participants at the 2012 Southwestern Finance Association annual meeting and the 2013
Southern Finance Association annual meeting (presented under the title: Employee Satisfaction and Firm Innovation) for helpful comments. All remaining
errors are our own.
∗ Corresponding author.
E-mail addresses: adhikarh@erau.edu (H.P. Adhikari), wonseok.choi01@utrgv.edu (W. Choi), sahn@uwgb.edu (N.B. Sah).

http://dx.doi.org/10.1016/j.jeconbus.2016.10.004
0148-6195/© 2016 Elsevier Inc. All rights reserved.
66 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76

components. Besides compensation, some factors which can lead to satisfied employees are treatment with care and respect,
long-term job protection and adequate retirement benefits, work-life balance, profit-sharing, etc. Several early theories
(Herzberg, Mausner, & Snyderman, 1959; Maslow, 1943; McGregor, 1960) ascribe stronger corporate performance (through
improved recruitment, retention and motivation) to employee satisfaction. Furthermore, Edmans (2011) finds that firms
with high levels of employee satisfaction generate superior returns for shareholders. Similarly, Faleye & Trahan (2011) show
that a labor-friendly corporate environment subsequently outperforms a size and industry-matched control sample in terms
of value creation, profitability, and productivity. However, it is conceivable that extremely benevolent corporate policies
favoring employees may cause them to shirk their duties. Bradley, Kim, & Tian (2016) find that unionization negatively
impacts corporate innovation. They argue that unionization provides adequate job protection but inadequate motivation
for promoting innovation. Cohen (2009) argues that union employees are less loyal to their firms. Also, one can argue that
excessive retirement benefits can shorten the employee work-horizon and may cause employees to slack off. Hence, the
popular view that all employee-friendly policies would pay off may not be true. So, if the manager of a firm wants to maximize
shareholder wealth then it may be beneficial to focus on some but not all employee friendly policies.
Most studies, simply consider a composite measure of employee satisfaction or analyze a stand-alone component in
isolation in order to study the impact of employee satisfaction on corporate outcomes. Hence, a holistic study which covers
the several different facets of employee satisfaction is warranted in order to help managers channelize their efforts in the
creation of a satisfied and productive workforce. We propose that employees will exert genuine effort for the firm if they are
satisfied with the firm. Thus, we test whether overall employee satisfaction in a firm has influence on its innovative ability.
Specifically, using multiple techniques, we establish that a firm’s employee-friendly polices as indicated by a composite
Employee Friendliness Index (EFI) escalates a firm’s future innovations and innovative efficiency. However, further analyses
of several stand-alone policies (components) indicate that they have differential impact on a firm’s future innovation output.
Specifically, our results suggest that cash profit sharing and employee involvement have a positive bearing on the innovative
output of a firm whereas union relationships can lead to a negative impact on innovative output of a firm. In addition, sub-
sample analysis reveals that employee friendly policies are more important in the industries having higher employee power
than the industries having lower employee power. Hence, managers need to be careful when they design employee-related
polices as one size doesn’t fit all. To the best of our knowledge, these research questions, in a comprehensive framework,
have not been empirically explored so far.
Our paper contributes to the emerging literature on finance and innovation by establishing that favorable corporate
environment (for employees) as measured by the Employee Friendliness Index (EFI) positively affects firm innovation and
innovative efficiency. These results taken along with the fact that firm innovation increases firm value (Hirshleifer et al.,
2013) suggest that employee friendly policies which lead to satisfied employees can increase firm value. Thus, we present a
novel channel through which satisfied employees can contribute towards increasing firm value − the channel of innovation.
Our study can lend valuable explanation to the results presented by Edmans (2011).
Our results also suggest that firms need to consistently create and maintain an employee friendly environment if they wish
to make their employees more innovative (as the impact of friendly policies diminishes over time). This idea is consistent
with several media reports which suggest that reward frequency is more important than size.1 Next, using subsamples, we
study labor intensive industries with high employee power (like manufacturing and wholesale) against other industries
with low employee power. We find that employee friendly policies play a more important role in industries having higher
employee power and lead to generation of more patents and citations. Finally, we also find that managers need to share
profits with their workforce and involve employees in strategic decision making process in order to nurture and increase
innovative output. At the same time, managers need to find alternate means (rather than relying on unions) of providing
adequate job protection to their employees in order to foster innovation, which is consistent with recent studies.
The remainder of the paper is organized as follows: Section 2 presents literature review and Section 3 develops hypotheses.
Section 4 describes data and presents descriptive statistics of the sample. Section 5 discusses our empirical results and
robustness tests. Section 6 concludes.

2. Literature review

Innovation is a knowledge producing activity which has long-term impact on a firm’s success and profitability. In the
ever-changing business environment, innovation is the key to success for the modern firms.2 There is also evidence in the
finance literature that innovation benefits a firm by increasing its value (Hirshleifer et al., 2013). Hence, several early studies
have tried to decipher the drivers of innovation.
However, it is also important to note that innovation is a long, tedious and risky process associated with huge costs
(Holmstrom, 1989). Hall & Lerner (2010) finds that R&D and other innovation investments are highly uncertain and the
result of success or failure is revealed over time. Given the nature of innovation, firms need to ensure that their employees
are sufficiently motivated to tread along the uncertain path. Hence, tolerance for failure in the short-term and reward for
success in the long-term would foster innovation (Manso, 2011). Using extensive data (for over 11,000 industrial scientists

1
See “How to increase productivity by employee happiness” by Martin Zwilling, Forbes, December 2, 2014.
2
See “Innovation: key to successful business” by Chad Brooks, Business News Daily, September 23, 2013.
 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76 67

and engineers), Sauermann & Cohen (2010) show that intrinsic motives (particularly the desire for intellectual challenge)
benefit innovation more than extrinsic motives (such as pay). Camerer & Hogarth (1999) report that non-pecuniary incentives
improve some features of individuals’ cognition such as memory, recall and simple problem-solving functions depending
upon the nature of the task at hand and the abilities of the individual. These findings suggest that success in innovation
(productivity) heavily depends on the internal satisfaction of the employees and not just compensation. Holmstrom (1989)
and Holmstrom & Milgrom (1991)’s multitasking framework shows that non-monetary incentive is not only a substitute for
pay, but a necessary tool for motivating innovation. They argue that non-monetary incentives must be used to encourage
innovation. Hence, firms need to focus on factors (besides compensation) which can lead to satisfied employees.
Early studies indicate a positive relationship between R&D expenditures and patents (Hausman, Hall, & Griliches, 1984;
Pakes & Griliches, 1980) and several others). These studies present the association between R&D expenditures and patent
generation by devising a patent production function. Over the years, extant literature has delved further into the drivers of
innovation. However, most studies explore the relationship between innovation and firm and market-related variables (like
firm size (Scherer, 1965), R&D expenditure (Acs & Audretsch, 1988; Hausman et al., 1984; Pakes & Griliches, 1980), industry
concentration (Levin, Cohen, & Mowery, 1985; Lunn, 1986), competition (Aghion & Howitt, 2005), corporate governance
(Meulbroek, Mitchell, Mulherin, Netter, & Poulsen, 1990), availability and type of financing (Benfratello, Schiantarelli, &
Sembenelli, 2008; Hsu, Tian, & Xu, 2014), bankruptcy laws (Acharya & Subramanian, 2009), and so on).
However, some recent studies attempt to connect innovation to employee-related aspects (Acharya, Baghai, &
Subramanian, 2014; Bradley et al., 2016). Along similar lines, we believe that employees play an important part in the pro-
cess of innovation. Alongside our study, a few contemporaneous studies are also investigating the link between employee
satisfaction and firm innovation (Chen, Chen, Hsu, & Podolski, 2016; Mao & Weathers, 2015; Mayer, Warr, & Zhao, 2015).
Although related, our work is distinct from these studies as we provide a holistic picture which covers several different
facets of employee satisfaction and helps managers channelize their efforts in the creation of a satisfied and productive
workforce. Our composite measure of employee friendliness consists of factors like union relations, cash profit sharing,
employee involvement, retirement benefits strength and work-life benefits. It is completely possible that all these factors
may not have a positive impact on innovation. Some corporate policies may actually cause employees to shirk their duties.
Bradley et al. (2016) find that unionization negatively impacts corporate innovation. They argue that unionization provides
adequate job protection but inadequate motivation for promoting innovation. Cohen (2009) argues that union employees
are less loyal to their firms. Also, one can argue that excessive retirement benefits can shorten the employee work-horizon
and may cause employees to slack off. Hence, we look at these factors individually as well.
Also, unlike other studies, we explore whether employee power influences the relationship between employee satisfac-
tion and innovation output. Employee power plays an important role in the success or failure of a firm (Balsmeier, Bermig, &
Dilger, 2013; Fauver & Fuerst, 2006 and several others). However, most of these studies use European data and corporations
for their analyses. In the U.S., it is possible that employee power varies from industry to industry (in the absence of any
regulations). Labor intensive industries, like manufacturing and wholesale, may have considerably more employee power
than other industries. Hence, we consider employee power as well in our study.

3. Hypothesis development

Our main research question for this study is: Can satisfied employees increase firm innovations and innovative efficiency and
thereby contribute towards increasing firm value?
As discussed earlier, we believe that employees play an important part in the process of innovation. The idea that
“employee treatment and satisfaction matters” has gained importance from both practitioners and academics. Alongside our
analysis, a few contemporaneous studies are also investigating the link between employee satisfaction and firm innovation.
Chen et al. (2016) primarily use a composite measure for employee treatment and find that firms which treat their employees
well have better innovative success. Another related paper, Mao & Weathers (2015) centers on providing evidence on the
relationship between employee treatment and success in innovation projects related to the core business of the firm. Mayer
et al. (2015) combine treatment of employees and diversity culture of the company (treatment of women and minorities) and
study its relationship with innovative efficiency. They find that better employee treatment and diversity policies increases
innovation efficiency.
Thus, certain non-pecuniary incentives can improve innovative capabilities of employees. One such benefit could be a
corporate environment which fosters friendliness between the firms and its employees.
So, our first hypothesis is:

H1a. Employee friendly firms innovate more.

Hypothesis 1a relates employee friendliness with the absolute measure of the innovation. This relationship, if positive,
would suggest that if a firm provides comfortable and appealing working environment, the employees will be innovative.
But, merely looking at this result, we cannot say that innovation as output (number of patents or citations) is cost effective
for the firm.
Hence, we want to further extend our analysis and test the link between the employee friendliness and the innovative
efficiency of the firm. By innovative efficiency, we mean a firm’s ability to generate patents or citations per dollar of sales.
68 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76

As the main objective of any firm is to maximize its value, firms would be more interested in measuring their innovative
efficiency than the absolute number of patents or citations.
Accordingly, our next hypothesis is:

H1b. Employee friendly firms have greater innovative efficiency.

Next, we want to explore the question that: Do all employee friendly policies work equally in all industries?
It is reasonable to assume that labor intensive industries will have high employee power and hence will function dif-
ferently from other industries. Hence, it will be interesting to study the role of employee power in the interplay between a
firm’s employee treatment and innovation output. Given the nature of the industries, we believe that employee treatment
will matter more in labor intensive industries (like manufacturing) as compared to other industries.
Accordingly, our second hypothesis is:

H2. Employee friendly policies have stronger influence on innovation output in industries having higher employee power.

Finally, we also want to study that: Should managers focus on some but not all employee friendly policies?
As discussed earlier, employee satisfaction is a very broad term and it can be considered as a mixture of several stand-alone
components. Hence, the use of a composite index may not be completely accurate to study the impact of employee friendly
policies. We believe that different stand-alone components of the composite index will contribute differently towards the
innovation output of a firm. This analysis can help managers focus on some (but not all) employee related policies in order
to boost the innovative output of their firms.
Accordingly, our third hypothesis is:

H3. Not all employee friendly policies create value for the firm.

4. Data

Our sample consists of all firms that are covered by the KLD Research and Analytics database across all industries. Next,
we merge the patents and citation database with the KLD database. We obtain patent and citation information from 2006
edition of the NBER Patent Citation database and use that database to construct firm innovation and innovation efficiency
variables. Hall, Jaffe, & Trajtenberg (2001) have a detailed explanation about the NBER Patent Citation database. We use the
KLD Research and Analytics database to construct Employee Friendliness Index (EFI). We collect accounting variables from
Compustat Industrial Annual Files and use them to construct control variables. We obtain stock related information from
CRSP. Our final sample that is used to examine the relationship between employee friendliness and firm innovation consists
of 12,183 firm year observations.

4.1. Employee Friendliness Index

We use KLD SOCRATES Research and Analytics database to construct Employee Friendliness Index. KLD database provides
ratings of a firm based on how the firm treats their employees. Employee Friendliness Index reflects different facets of human
resource practices within the firm including: employee involvement in the firm, union relationship, cash profit sharing, and
work/life and retirement benefits. KLD rates companies using data from various sources such as company filings, public
media, and government data. They provide rating of firms in seven major areas: Community, corporate governance, diversity,
employee relations, environment, human rights, product safety and quality. As described in Ertugrul (2013), KLD rates firms
based on screens called “strengths” and “concerns”. We are interested in the screens related to employee treatment. KLD
describes these screens as:

1) Union relations: Whether the company has taken exceptional steps to treat its unionized work force fairly.
2) Cash profit sharing: Whether the company has a cash profit-sharing program through which it has recently made
distributions to a majority of its workforce.
3) Employee involvement: Whether the company strongly encourages worker involvement and/or ownership through stock
options available to a majority of its employees; gain sharing, stock ownership, sharing of financial information, or
participation in management decision making.
4) Retirement benefits strength: Whether the company has a notably strong retirement benefits program.
5) Work/life benefits: Whether the company has outstanding employee benefits or other programs addressing work/family
concerns, (e.g., childcare, elder care, or flextime).

KLD assigns a 0/1 rating for each of these categories. To construct the Employee Friendliness index (EFI), we sum the rating
scores for each category and create an index that ranges from zero to five. A higher value would indicate a more employee
friendly firm. By construction, we measure an annual EFI score for an individual firm. This variable has cross-sectional as
well as time-varying components and hence, there are deviations across years and firms. Also, in order to mitigate the
problem associated with the possible lagged reaction of patents and citations to changes in EFI, we study the impact of EFI
on innovation productivity with a lag of one, two and three years.
 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76 69

4.2. Firm innovation and innovative efficiency

We obtain patent and citation information from 2006 edition of the NBER Patent Citation database and use that database
to construct Firm Innovation and Innovative Efficiency variables. This database provides detailed information about every
patent granted by its regulatory body: United states patent and Trade Office (USPTO). The data ranges from 1976 to 2006. With
this database and following Sevilir & Tian (2012), we construct two measures: Firm Innovation and Innovation Efficiency

1) Firm innovation: The first measure of firm innovation is the number of patents filed (and eventually granted in some future
time) by a firm in a year. Following Griliches, Pakes, & Hall (1988) and Sevilir & Tian (2012), we use patent’s application year
instead of its grant year because the application year better matches the timing of innovation. Sevilir & Tian (2012) argue
that simple count of patents may not differentiate impactful innovations from incremental technological discoveries.
So, we construct the second measure of firm innovation by counting the number of non-self-citations received by each
patent of the firm in a given year. The number of patents and citations capture the innovation output of a firm.
2) Innovative efficiency: As discussed earlier, firms main interest lies in maximizing their value. So, firms are more interested
in their innovative efficiency rather than the number of patents or citations. Hence, to measure innovative efficiency, we
use the logarithm of the ratio of the number of patents to sales and the logarithm of the ratio of number of patent citations
to sales.3

4.3. Control variables

We follow the existing innovation literature to obtain the control variables. We control for a set of firm characteristics
that may affect firm innovation or innovative efficiency. These control variables are free cash flow, firm size (the natural
logarithm of sales), Market-to-Book Ratio (growth opportunities), leverage, R&D expenditure scaled by total assets, firm age,
tangibility (property, plant and equipment scaled by book value of assets) and sales growth. Several contemporary studies
control for these standard set of variables as they are known to impact innovation output.
We use unbalanced panel data in this study wherein our main variables of interest are patents and citations count data.
The distribution of these counts is discrete and is limited to non-negative values. As such, the patents and citations count data
(or a simple transformation of it) may not follow a normal distribution and hence the use of ordinary OLS may produce biased
and inefficient estimates. It would be more appropriate to consider that the patents and citations count data has a Poisson
distribution. Negative binomial regression analysis is a standard method used to model data with Poisson distribution.
Taking a cue from existing literature (Hausman et al., 1984), we use the negative binomial count data models (with year and
industry fixed effects and clustered robust-variance estimation) in our analyses as it is better suited for patent and citation
data (count data) as our primary tool for our study.

4.4. Descriptive statistics

We present descriptive statistics in Table 1. It reports the number of firm year observations (N), mean, median and
standard deviation of each variable. It also reports the 1st and 99th percentiles. The mean (median) EFI is 0.364 (0). The
mean (median) number of patents is 17.963 (0) and mean number of citation is 128.295 (0).
We start our empirical investigation with the univariate test designed to observe whether the firm characteristics are
significantly different for firms with high employee satisfaction (High Employee Friendliness Index) and the firms with low
employee satisfaction (Low Friendliness Index). Table 2 illustrates that among 12,183 firm year observations 3,319 firm year
observations (around 27%) have High Employee Friendliness Index and 8,864 (around 73%) have Low Friendliness Index.
High Employee Friendliness Index group has an average index of 1.335 which is significantly different at 1% level from Low
Friendliness Index group which has an average of 0.000 from construction. Groups with high and low friendliness indexes
differ significantly at 1% level in all the firm characteristics except leverage.

5. Empirical results

5.1. Multivariate tests

The general specification for our multivariate tests can be expressed as:

Innovation Outputi,t+1 = ˇ0 + ˇ1 EFIi,t + ˇ2 ln (Sales)i,t + ˇ3 ln(MB)i,t + ˇ4 FreeCashi,t + ˇ5 Leveragei,t + ˇ6 R&D/Asseti,t

+ ˇ7 PPE/Asseti,t + ˇ8 ln (FirmAge)i,t +ˇ9 Sales Growthi,t + Year&IndustryFixedEffects + εi,t

3
Our approach of scaling patents and citations by sales is similar to Ciftci, Lev, & Radhakrishnan (2011). Studies have also used the number of patents
divided by R&D expenditure (Hirshleifer, Hsu, & Li, 2013) to measure innovation efficiency. However, due to missing and zero values for R&D expenditures,
we feel that our measure is appropriate for this study.
70 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76

Table 1
Descriptive statistics.
Table 1 presents descriptive statistics of the sample firms. The sample period ranges from 1991 to 2004. Employee Friendliness Index is assigned a rating
of one, two, three, four, or five for each category. a and b denote a minimum value and a maximum value respectively. Detailed definitions of variables are
in Appendix A.

Variable Names N Mean Median 1 percentile 99 percentile

EFI 12,183 0.364 0.000 0.0 a 3.0 b

Union Relationship 12,183 0.008 0.000 0.0 a 1.0 b
Cash Profit Sharing 12,183 0.100 0.000 0.0 a 1.0 b
Employee Involvement 12,183 0.119 0.000 0.0 a 1.0 b
Retirement Benefit Strength 12,183 0.051 0.000 0.0 a 1.0 b
Work/Life Benefit 12,183 0.086 0.000 0.0 a 1.0 b
Number of Patents 12,183 17.963 0.000 0.000 354.000
Number of Citations 12,183 128.295 0.000 0.000 2,289.000
Number of Citations (Weighted) 12,183 290.769 0.000 0.000 5,206.600
Sales 12,174 4,645.890 1,397.060 11.053 56,042.000
MB 11,967 3.318 2.343 0.627 20.291
Free Cash 11,769 0.070 0.075 −0.356 0.285
Leverage 12,135 0.232 0.214 0.000 0.875
R&D/Asset 12,183 0.029 0.000 0.000 0.303
PPE/Asset 11,779 0.280 0.223 0.001 0.890
Firm Age 12,183 25.049 22.000 3.000 54.000

Table 2
Univariate tests.
Table 2 presents univariate results indicating differences in sample characteristics by groups of Employee Friendliness Index. High Employee Friendliness
Index group consists of firms with rating greater than zero and Low Employee Friendliness Index group represents firms having zero rating. Detailed
definitions of variables are in Appendix A.

H L Differences (H–L) P-value

Observations 3,319 8,864

EFI 1.335 0.000 1.335 (<0.001)
N of Patents 48.076 6.688 41.388 (<0.001)
N of Citations 346.022 46.769 299.253 (<0.001)
N of Citations (Weighted) 802.303 99.232 703.072 (<0.001)
Sales 8,898.810 3,051.830 5,846.980 (<0.001)
MB 3.777 3.144 0.633 (<0.001)
Free Cash 0.083 0.065 0.018 (<0.001)
Leverage 0.229 0.233 −0.005 0.216
R&D/Asset 0.033 0.027 0.005 (<0.001)
PPE/Asset 0.313 0.267 0.046 (<0.001)
Firm Age 30.166 23.133 7.033 (<0.001)

5.1.1. Firm innovation

Table 3 presents the results of the negative binomial count data models (with year and industry fixed effects and clustered
robust-variance estimation) showing the relationship between Employee Friendliness Index and the firm innovation. We
have the number of patents (count) as the dependent variables in the first three models. Model 1 uses the number of patents
in the next year as the dependent variable (representing a firm’s innovation productivity). The coefficient estimate of EFI is
positive and significant at 1% level suggesting that firms maintaining higher employee satisfaction obtain a larger number of
patents, that is, they are more innovative. Specifically, if a firm were to increase their EFI score by one point, the difference
in the logs of expected counts of patents would increase by 0.203 next year (or the firm would produce 1.23 more patents),
while holding the other variables in the model constant. In Models 2 and 3, we replace dependent variable by the number of
patents in two subsequent years and three subsequent years after measuring EFI respectively. The coefficient estimates of
EFI continue to be positive and significant at 1% level (coefficients of 0.194 and 0.175 are significant). The coefficients of EFI
in Models 2 and 3 show that maintaining employee satisfaction now will have significant impact in second and third years
also but the impact decreases with time. First three models in this table also indicate that sales, MB ratio, free cash flow,
R&D expenditure scaled by asset and firm age all have significant positive effect on the number of patents filed in all three
specifications.
In Models 4-6, the dependent variable is the number of patent citations (count). We use the number of patent citations
as another measure of firm innovativeness. Model 4 uses the number of patent citations in the next year as the dependent
variable. The coefficient estimate of EFI is (once again) positive and significant at 1% level suggesting that firms with satisfied
employees obtain a larger number of citations, another sign of being more innovative (increasing EFI score one unit in a
year would lead to an increase in the log of count of citations by 0.195 (or 1.22 more citations) in the next year. In Models
5 and 6, we replace dependent variables by the number of patent citations in two subsequent years and three subsequent
 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76 71

Table 3
Employee friendliness index and innovation: using negative binomial count data models.
Table 3 reports results regarding the relationship between Employee Friendliness Index and Innovation using negative binomial count data models (with
year and industry fixed effects and clustered robust-variance estimation). Dependent variables are the number of patents and citations that a firm applies
for in a given year. Windows for dependent variables ranges from year t+1 to year t+3. Detailed definitions of other variables are in Appendix A. Numbers
in parentheses are heteroskedascity robust p-values. All models include year and industry dummies. Industry dummies are defined using Fama and French
12 industry classifications.

(1) (2) (3) (4) (5) (6)

PATt+1 PATt+2 PATt+3 Citet+1 Citet+2 Citet+3

EFI 0.203 0.194 0.175 0.195 0.191 0.176

(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Sales 0.593 0.616 0.640 0.626 0.599 0.607
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
ln(MB) 0.271 0.171 0.164 0.137 0.126 0.161
(<0.001) (<0.001) (<0.001) (0.012) (0.044) (0.018)
Free Cash 0.932 1.081 0.154 0.892 0.774 −0.581
(0.013) (0.023) (0.790) (0.091) (0.278) (0.459)
Leverage −0.382 −0.197 −0.062 −0.287 −0.294 −0.392
(0.025) (0.313) (0.780) (0.224) (0.284) (0.190)
R&D/Asset 14.907 15.758 15.484 13.309 15.605 15.218
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
PPE/Asset −0.368 −0.386 −0.370 −0.703 −0.655 −0.395
(0.013) (0.019) (0.044) (<0.001) (0.004) (0.120)
ln(Firm Age) 0.292 0.297 0.243 0.226 0.324 0.341
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Sales Growth −0.056 0.054 0.087 −0.099 −0.171 −0.168
(0.546) (0.668) (0.566) (0.477) (0.351) (0.400)
Intercept −8.635 −8.883 −8.732 −12.243 −12.634 −12.632
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Year/Industry YES YES YES YES YES YES
N 11,364 8,224 5,862 11,364 8,224 5,862
LR (all coefficients=0) 549,761 495,957 447,831 4,611,790 3,888,607 3,226,886

years after measuring EFI respectively. The coefficient estimates of EFI continue to be positive and significant at 1% level.
The coefficients of EFI in Models 5 and 6 shows that maintaining employee satisfaction will have a positive and significant
impact on patent citations in second and third years also (coefficients of 0.191 and 0.176 are significant) but (similar as in
case of number of patents) the impact of EFI on number of citations decreases with time. In all of the above specifications,
we control for year and industry fixed effects and the control variables have expected signs.

5.1.2. Innovation efficiency

In Table 4, we analyze the relationship between employee satisfaction and innovative efficiency of the firm. We measure
employee satisfaction by the composite Employee Friendliness Index (EFI) and Innovative efficiency by the logarithm of the
number of patents per unit sales in the first three models (Models 1 and 3) and by the logarithm of the number of citations
per unit of sales in the next three models (Models 2 and 4). Models 1 and 2 use the OLS methodology whereas Models 3 and
4 use the Generalized Method of Moments (GMM) methodology.
In Models 1 and 2 where the dependent variable is the logarithm of the ratio of number of patents to sales revenue next
year, the coefficient estimate of EFI is positive and significant at 1% level. This result shows that the more employee friendly
the firm is, the higher its innovative efficiency in the next year. In all these regressions, we control for a comprehensive set
of firm characteristics that affect firm’s innovative efficiency and the control variables have expected signs.
Models 3 and 4 have the logarithm of the number of patent citations per unit sales revenue next year as the dependent
variable. The coefficient estimate of EFI is again positive and significant at 1% level. This result further establishes that the
more employee-friendly the firm is, the higher its innovative efficiency is in the next year. Here also, we control for a
comprehensive set of firm characteristics that affect firm’s innovative efficiency and the control variables have expected
signs.
Using two different estimation techniques provides robustness to our results. Thus, we find support for our first hypothesis
and we can say that firms with congenial work environments innovate more and have greater innovative efficiency.

5.1.3. Role of employee power

In Table 5, we present the results related to the role of employee power in the interplay between a firm’s employee
treatment and innovation output. We find that employee treatment matters more in labor intensive industries (covering
Standard Industrial Classification (SIC) numbers from 2000 to 5999) as compared to other industries.
We find that better employee treatment results in more patents as well as citations in labor intensive industries. So, it
can be said that employee friendliness and congenial work environment can lead to higher innovation output in industries
with high employee power.
72 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76

Table 4
Employee friendliness index and innovation efficiency.
Table 6 provides results showing the relationship between the Employee Friendliness Index and the Firm Innovation Efficiency. Innovation efficiencies
are measured by the number of patents scaled by sales and the number of citation scaled by sales. Models (1) and (2) use the OLS methodology whereas
Models (3) and (4) use the GMM methodology. As the dependent variables are not counts of patents and citations, we use the natural logarithm of patents
and citations scales by sales in these tests. Detailed definitions of other variables are in Appendix A. Numbers in parentheses are heteroskedascity robust
p-values. Model (1) and (2) include year and industry dummies. Industry dummies are defined using Fama and French 12 industry classifications.

(1) (2) (3) (4)

log (PAT/Sales)t+1 log (Cite/Sales)t+1 log (PAT/Sales)t+1 log (Cite/Sales)t+1

EFI 0.001 0.004 0.001 0.007

(0.001) (<0.001) (<0.001) (<0.001)
Sales −0.001 0.001 0.000 0.002
(<0.001) (0.057) (0.003) (<0.001)
ln(MB) 0.000 −0.006 −0.001 −0.009
(0.830) (<0.001) (0.007) (<0.001)
Free Cash 0.005 0.131 0.015 0.169
(0.070) (<0.001) (<0.001) (<0.001)
Leverage 0.000 0.006 0.000 −0.002
(0.931) (0.083) (0.642) (0.455)
R&D/Asset 0.081 0.289 0.123 0.450
(<0.001) (<0.001) (<0.001) (<0.001)
PPE/Asset 0.005 0.017 0.002 0.004
(<0.001) (<0.001) (0.001) (0.062)
ln(Firm Age) 0.000 −0.001 0.001 0.006
(0.157) (0.099) (<0.001) (<0.001)
Sales Growth 0.000 0.002 −0.001 −0.003
(0.879) (0.389) (0.232) (0.186)
Intercept 0.009 0.050 0.000 −0.031
(<0.001) (<0.001) (0.857) (<0.001)
Year/Industry YES YES
N 11,364 11,364 11,364 11,364
Adj R2 25.82 25.58

For GMM: Instruments used are EFI, ln(Sales), ln(MB), Free Cash, Leverage, R&D/Asset, PPE/Asset, ln(Firm Age), and Sales Growth. The Sargan test indicates
that the instruments used in the GMM estimation are valid and there is no over-identifying issue.

Table 5
Employee friendliness index and innovation: employee power high vs. low.
Table 5 reports results regarding the relationship between Employee Friendliness Index and Innovation in industries having high employee power vs.
industries having lower employee power by executing negative binomial count data models (with year and industry fixed effects and clustered robust-
variance estimation). Dependent variables are the number of patents and citations that a firm applies for in a given year. High Employee Power industries
are defined as the Manufacturing and Wholesale industry (SIC 2000s, 3000s, and 5000s) and other industries are considered as Low Employee Power.
Detailed definitions of other variables are in Appendix A. Numbers in parentheses are heteroskedascity robust p-values. All models include year and
industry dummies. Industry dummies are defined using Fama and French 12 industry classifications.

(1) (2) (3) (4)

High Low High Low

PATt+1 Citet+1

EFI 0.177 0.132 0.198 0.154

(<0.001) (0.147) (<0.001) (0.233)
Sales 0.612 0.728 0.541 1.096
(<0.001) (<0.001) (<0.001) (<0.001)
ln(MB) 0.087 0.931 −0.020 0.834
(0.036) (<0.001) (0.734) (<0.001)
Free Cash 0.625 0.800 2.071 −1.527
(0.115) (0.458) (<0.001) (0.329)
Leverage −0.287 −1.593 −0.336 0.446
(0.123) (<0.001) (0.204) (0.476)
R&D/Asset 10.884 27.464 12.112 13.328
(<0.001) (<0.001) (<0.001) (<0.001)
PPE/Asset −0.372 −0.582 −0.516 −2.168
(0.033) (0.122) (0.025) (<0.001)
ln(Firm Age) 0.160 0.400 0.161 −0.064
(<0.001) (0.002) (0.013) (0.717)
Sales Growth −0.087 −0.075 −0.314 −0.158
(0.387) (0.775) (0.034) (0.703)
Intercept −7.105 −10.234 −10.034 −15.831
(<0.001) (<0.001) (<0.001) (<0.001)
Year/Industry YES YES YES YES
N 6,459 4,905 6,459 4,905
LR (all coefficients=0) 483,467 66,799 4,028,890 583,273
 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76 73

Table 6
Innovation and the components of employee friendliness index.
Table 6 presents results regarding the relationship between the components of Employee Friendliness Index and Innovation using negative binomial count
data models (with year and industry fixed effects and clustered robust-variance estimation). Dependent variables are the number of patents and citations
that a firm applies for in a given year. Detailed definitions of other variables are in Appendix A. Numbers in parentheses are heteroskedascity robust
p-values. All models include year and industry dummies. Industry dummies are defined using Fama and French 12 industry classifications.

(1) (2)

PATt+1 Citet+1

Union Relationship -0.218 -0.455

(0.017) (<0.001)
Cash Profit Sharing 0.962 0.974
(<0.001) (<0.001)
Employee Involvement 0.600 0.755
(<0.001) (<0.001)
Retirement Benefit Strength 0.028 -0.020
(0.692) (0.829)
Work/Life Benefit -0.033 0.027
(0.786) (0.866)
Sales 0.633 0.668
(<0.001) (<0.001)
ln(MB) 0.207 0.156
(<0.001) (0.028)
Free Cash 2.998 3.981
(<0.001) (0.001)
Leverage 0.118 −0.031
(0.637) (0.935)
R&D/Asset 32.455 32.674
(<0.001) (<0.001)
PPE/Asset −0.520 −0.552
(0.001) (0.025)
ln(Firm Age) 0.451 0.331
(<0.001) (<0.001)
Sales Growth −0.253 −0.403
(0.136) (0.083)
Intercept −5.306 −2.673
(<0.001) (<0.001)
Year/Industry YES YES
N 11,364 11,364
LR (all coefficients=0) −20,670 −20,447

5.1.4. Impact of individual components of EFI

Finally, in order to present a holistic analysis and to help managers channelize their efforts in the creation of a satisfied
and productive workforce, we analyze the individual components using the negative binomial count data models (with year
and industry fixed effects and clustered robust-variance estimation). Specifically, we analyze union relations, cash profit
sharing, employee involvement, retirement benefits strength and work-life benefits.
We present our analysis in Table 6. Our results suggest that cash profit sharing and employee involvement have a positive
bearing on the innovative output of a firm. These results are in line with several academic studies (Miller & Monge, 1986)
as well as reports in the financial press.4 Also, we find that union relationships can lead to a negative impact on innovative
output of a firm. Thus, our results are similar to the findings of Bradley et al. (2016) who also find that find that unionization
negatively impacts on corporate innovation.
Based on our results we find support for our third hypothesis that not all employee friendly policies create value for
the firm. We can also suggest that managers need to share profits with their workforce and involve employees in strategic
decision making process in order to nurture and increase innovative output. At the same time, managers need to find
alternate means (rather than relying on unions) of providing adequate job protection to their employees in order to foster
innovation.

5.1.5. Robustness tests

5.1.5.1. Using alternative estimation techniques. Several recent studies point towards the causal impact of employee-related
policies on innovation productivity (Bradley et al., 2016). However, one can easily argue that more innovative firms (e.g.:
software and technology firms) treat their employees better. Hence, our specifications could suffer from potential problems
related to reverse causality. Also, there could be omitted variables in our specifications. These potential issues of causality,
omitted variables and endogeneity could impact the validity of our results. We remedy this problem by using various
alternative estimation techniques. We execute Ordinary Least Squares (OLS) regressions (Models 1 and 5) with year and

4
See “Company decision making doesn’t need to start at the top” by Robert Jordan, Forbes, October 27, 2012.
74 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76

industry level controls, Weighted Least Squares (WLS) regressions (Models 2 and 6), simultaneous equations model using
the Three-Stage Least Squares (3SLS) procedure (Models 3 and 7) (which addresses the potential endogeneity and causality
problem between patent production and employee friendliness by jointly estimating both in a simultaneous equations
framework) and GMM procedure (Models 4 and 8) to resolve our empirical problems.5 Although we have tried to remedy
potential problems related to causality, omitted variables and endogeneity, the results have to be treated with caution. Our
results are reported in Table B1 of Appendix B.
The dependent variable in the first four models is the logarithm of number of patents in next year, whereas the depen-
dent variable in the last four models is the logarithm of number of patent citations in next year. We control for the same
comprehensive set of variables that may have influence on firm innovation and we get the similar results. These results are
consistent with the results we find above in our baseline regression models using negative binomial count data models.
The coefficient estimate of EFI is positive and significant at 1% level in all the alternate specification techniques. Overall,
employee friendly corporate environment improves innovation productivity.

5.1.6. Using dummy variable for EFI

EFI is measured as an ordered categorical. The use of the ordered categorical variable shows the summated effect of
Employee Friendliness Index. In order to address the concern that there might be the regime change(s) which deviate from
the marginal effects of employee friendly policies, we create several thresholds and use them instead of directly using EFI.
Specifically, EFI > 0 dummy is encoded as 1 if EFI is greater than 0 and 0 otherwise, EFI > 1 dummy is encoded as 1 if EFI is
greater than 1 and 0 otherwise and EFI > 2 dummy is encoded as 1 if EFI is greater than 2 and 0 otherwise.6
Similar to our baseline regressions, our analysis using dummy variables (results are reported in Table B2 of Appendix B)
also reveals that firms with friendly work environments innovate more. Our results also suggest an interesting pattern, firms
which have an EFI score of greater than 2 feel the most positive impact on their patent and citation production by creating
a friendly workplace.

6. Conclusion

Innovation is one of the most important aspects for the growth of modern corporate firms. Prior literature viewed
corporate innovation as a function of firm and market-related variables. However, recent studies have indicated that labor
related issues can impact corporate innovation. Our study is in similar vein. We believe that employees play an important part
in the process of innovation and hence firms must make continued efforts towards creating and maintaining a friendly work
environment which increases employee satisfaction. Through multiple and robust tests, we find that a congenial corporate
environment measured using Employee Friendliness Index (EFI) positively affects firm innovation and innovative efficiency.
We also find that labor intensive industries are more sensitive to employee friendly policies as compared to other industries.
Finally, our results also indicate that managers need to focus on some but not all employee related policies in order to boost
the innovative output of their firms.
As firm innovation increases firm value, employee friendly policies which lead to satisfied employees can increase firm
value. So, our finding suggests that a firm’s employee-friendly policies that motivate innovation could be a channel leading
to higher firm value. Our study can also help explain the results presented by Edmans (2011) that employee satisfaction
can increase shareholder returns. Our results also suggest that firms need to consistently create and maintain an employee
friendly environment if they wish to make their employees more innovative.

Appendix A. Definitions of variables

This table contains the definitions of variables.

Variable Names Definitions of variables

Employee Friendliness Index (EFI) Employee Friendliness Index created using data from KLD SOCRATES database.
Number of Patents A number of applications for patents given a firm-year
Number of Citations A number of citations for patents given a firm-year
Number of Citations (Weighted) Citations for patents adjusted by time
MB Market to book ratio defined as the market value of equity to the book value of equity.
Free Cash Free cash flow defined as [(Operating Cash Flow-Capital Expenditure)/Total Assets]
Sales Logarithm of the sales
Leverage Debt in current liabilities plus long-term debt divided by total assets
R&D/Asset Research and development expenses scaled by total assets
PPE/Asset Net property, plants and equipment scaled by total assets
Firm Age Firm age is counted from the year the firm appeared in the Compustat
Sales Growth Annual percentage increase in sales

5
Additional details about these alternative estimation techniques are provided in Appendix B.
6
Additional dummy variables for higher levels of EFI are not needed due to lack of observations.
 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76 75

Appendix B. (Robustness tests)

Table B1 provides regression results using various alternative estimation methods Models (1)–(4) examine the patents
as dependent variable whereas Models (5)–(8) examine the citations. Rather than using the counts of patents and citations
we use the natural logarithm of patents and citations in these tests. Detailed definitions of other variables are in Appendix
A. Numbers in parentheses are heteroskedascity robust p-values. All models include year and industry dummies (except
GMM). Industry dummies are defined using Fama and French 12 industry classifications.
Table B2 provides results regarding the relationship between the level of Employee Friendliness Index and Innovation
using threshold dummy variables by executing negative binomial count data models (with year and industry fixed effects
and clustered robust-variance estimation). Dependent variables are the number of patents and citations that a firm applies
for in a given year. EFI > 0 dummy is encoded as 1 if EFI > 0 and 0 otherwise. EFI > 1 dummy is encoded as 1 if EFI > 1 and 0
otherwise. EFI > 2 dummy is encoded as 1 if EFI > 2 and 0 otherwise. Detailed definitions of other variables are in Appendix
A. Numbers in parentheses are heteroskedascity robust p-values. All models include year and industry dummies. Industry
dummies are defined using Fama and French 12 industry classifications.

Table B1
Employee friendliness index and innovation: using alternative estimation methods.

(1) (2) (3) (4) (5) (6) (7) (8)

log(PAT)t+1 log(Cite)t+1

OLS WLS 3SLS GMM OLS WLS 3SLS GMM

EFI 0.175 0.261 0.206 0.268 0.156 0.315 0.219 0.284

(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Sales 0.242 0.220 0.236 0.229 0.238 0.260 0.262 0.264
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
ln(MB) 0.055 0.027 0.076 0.027 0.011 0.000 0.002 −0.014
(0.002) (0.137) (0.009) (0.170) (0.638) (0.997) (0.949) (0.610)
Free Cash 1.023 2.247 2.080 2.627 2.020 3.066 3.566 3.682
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Leverage −0.073 0.034 0.326 0.103 0.050 −0.091 0.581 0.014
(0.274) (0.618) (0.007) (0.122) (0.560) (0.339) (<0.001) (0.870)
R&D/Asset 4.482 10.265 18.930 11.569 5.344 11.280 21.263 12.972
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
PPE/Asset 0.242 −0.209 −0.296 −0.181 0.311 −0.002 −0.478 0.086
(0.002) (<0.001) (<0.001) (<0.001) (0.004) (0.982) (<0.001) (0.168)
ln(Firm Age) 0.070 0.256 0.365 0.272 0.039 0.365 0.349 0.369
(<0.001) (<0.001) (<0.001) (<0.001) (0.088) (<0.001) (<0.001) (<0.001)
Sales Growth −0.166 −0.269 −0.392 −0.292 −0.183 −0.345 −0.464 −0.372
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Intercept −1.076 −2.056 −2.390 −2.251 0.110 −2.651 −1.585 −2.807
(<0.001) (<0.001) ( < .001) (<0.001) 0.423 (<0.001) (<0.001) (<0.001)
Year/Industry YES YES YES YES YES YES
N 11,364 11,364 11,364 11,364 11,364 11,364 11,364 11,364
Adj R2 48.56 25.05 37.27 47.68 19.94 39.91

For 3SLS: We set up the following simultaneous equations: Innovationt+1 = f (EFIt , Controlst , Zt ).
EFIt = f (Innovationt , EFIt−1 , EFIt−2 ).
All control variables in the first equation would be considered exogenous. Whereas, EFI is estimated as a function of the firm innovation and the past values
of EFI.
For GMM: Instruments used are EFI, ln(Sales), ln(MB), Free Cash, Leverage, R&D/Asset, PPE/Asset, ln(Firm Age), and Sales Growth. The Sargan test indicates
that the instruments used in the GMM estimation are valid and there is no over-identifying issue.

Table B2
Innovation and the level of Employee Friendliness Index: Using dummy variables for EFI.

(1) (2) (3) (4) (5) (6)

PATt+1 Citet+1 PATt+1 Citet+1 PATt+1 Citet+1

EFI > 0 0.219 0.228

(<0.001) (0.007)
EFI > 1 0.270 0.171
(0.002) (0.129)
EFI > 2 0.773 0.632
(<0.001) (<0.001)
Sales 0.637 0.656 0.639 0.666 0.641 0.664
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
ln(MB) 0.188 0.140 0.192 0.150 0.197 0.157
(0.001) (0.049) (<0.001) (0.034) (<0.001) (0.028)
76 H.P. Adhikari et al. / Journal of Economics and Business 90 (2017) 65–76

Table B2 (Continued)

(1) (2) (3) (4) (5) (6)

PATt+1 Citet+1 PATt+1 Citet+1 PATt+1 Citet+1

Free Cash 3.132 4.177 3.163 4.136 3.139 4.175

(<0.001) (0.001) (<0.001) (0.001) (<0.001) (0.001)
Leverage 0.249 0.138 0.250 0.126 0.191 0.073
(0.312) (0.712) (0.308) (0.736) (0.438) (0.847)
R&D/Asset 33.270 34.260 33.347 34.531 33.476 34.607
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
PPE/Asset −0.477 −0.362 −0.394 −0.208 −0.371 −0.214
(0.003) (0.197) (0.018) (0.490) (0.025) (0.482)
ln(Firm Age) 0.425 0.308 0.422 0.302 0.415 0.296
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Sales Growth −0.251 −0.375 −0.240 −0.371 −0.232 −0.370
(0.133) (0.111) (0.147) (0.119) (0.158) (0.116)
Intercept −5.309 −2.681 −5.309 −2.765 −5.298 −2.725
(<0.001) (<0.001) (<0.001) (<0.001) (<0.001) (<0.001)
Year/Industry YES YES YES YES YES YES
N 11,364 11,364 11,364 11,364 11,364 11,364
LR (all coefficients = 0) 549,755 4,611,794 549,745 4,611,781 549,759 4,611,785

References

Acharya, V. V., & Subramanian, K. V. (2009). Bankruptcy codes and innovation. Review of Financial Studies, 22, 4949–4988.
Acharya, V. V., Baghai, R. P., & Subramanian, K. V. (2014). Wrongful discharge laws and innovation. Review of Financial Studies, 27, 301–346.
Acs, Z. J., & Audretsch, D. B. (1988). Innovation in large and small firms: an empirical analysis. American Economic Review, 678–690.
Aghion, P., & Howitt, P. (2005). Growth with quality-improving innovations: an integrated framework. Handbook of Economic Growth, 1, 67–110.
Balsmeier, B., Bermig, A., & Dilger, A. (2013). Corporate governance and employee power in the boardroom: An applied game theoretic analysis. Journal of
Economic Behavior & Organization, 91, 51–74.
Benfratello, L., Schiantarelli, F., & Sembenelli, A. (2008). Banks and innovation: Microeconometric evidence on Italian firms. Journal of Financial Economics,
90(2), 197–217.
Bradley, D. J., Kim, I., & Tian, X. (2016). The causal effect of labor unions on innovation. Management Science, forthcoming.
Camerer, C. F., & Hogarth, R. M. (1999). The effects of financial incentives in experiments: A review and capital-labor-production framework. Journal of
Risk and Uncertainty, 19(1–3), 7–42.
Chen, C., Chen, Y., Hsu, P. H., & Podolski, E. J. (2016). Be nice to your innovators: Employee treatment and corporate innovation performance. Journal of
Corporate Finance, 39, 78–98.
Ciftci, M., Lev, B., & Radhakrishnan, S. (2011). Is research and development mispriced or properly risk adjusted? Journal of Accounting, Auditing & Finance,
26(1), 81–116.
Cohen, L. (2009). Loyalty-based portfolio choice. Review of Financial Studies, 22(3), 1213–1245.
Edmans, A. (2011). Does the stock market fully value intangibles? Employee satisfaction and equity prices. Journal of Financial Economics, 101(3), 621–640.
Ertugrul, M. (2013). Employee-friendly acquirers and acquisition performance. Journal of Financial Research, 36(3), 347–370.
Faleye, O., & Trahan, E. A. (2011). Labor-friendly corporate practices: Is what is good for employees good for shareholders? Journal of Business Ethics,
101(1), 1–27.
Fauver, L., & Fuerst, M. E. (2006). Does good corporate governance include employee representation? Evidence from German corporate boards. Journal of
Financial Economics, 82(3), 673–710.
Griliches, Z., Pakes, A., & Hall, B. H. (1988). The value of patents as indicators of inventive activity (No. w2083). National Bureau of Economic Research.
Hall, B. H., Jaffe, A. B., & Trajtenberg, M. (2001). The NBER patent citation data file: Lessons, insights and methodological tools (No. w8498). National Bureau of
Economic Research.
Hall, B. H., & Lerner, J. (2010). The financing of R&D and innovation. Handbook of the Economics of Innovation, 1, 609–639.
Hausman, J. A., Hall, B. H., & Griliches, Z. (1984). Econometric models for count data with an application to the patents-R&D relationship. Econometrica,
52(4), 909–938.
Herzberg, F., Mausner, B., & Snyderman, B. (1959). The motivation to work. New York: John Wiley & Sons.
Hirshleifer, D., Hsu, P. H., & Li, D. (2013). Innovative efficiency and stock returns. Journal of Financial Economics, 107(3), 632–654.
Holmstrom, B. (1989). Agency costs and innovation? Journal of Economic Behavior & Organization, 12(3), 305–327.
Holmstrom, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership, and job design. Journal of Law, Economics, &
Organization, 24–52.
Hsu, P. H., Tian, X., & Xu, Y. (2014). Financial development and innovation: Cross-country evidence. Journal of Financial Economics, 112(1), 116–135.
Levin, R. C., Cohen, W. M., & Mowery, D. C. (1985). R & D appropriability, opportunity, and market structure: new evidence on some Schumpeterian
hypotheses. American Economic Review, 20–24.
Lunn, J. (1986). An empirical analysis of process and product patenting: a simultaneous equation framework. The Journal of Industrial Economics, 319–330.
Manso, G. (2011). Motivating innovation. Journal of Finance, 66(5), 1823–1860.
Mao, C. X., & Weathers, J. (2015). Employee treatment and firm innovation.. Available at SSRN (http://ssrn.com/abstract=2648588)
Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 50(4), 370.
Mayer, R. C., Warr, R. S., & Zhao, J. (2015). Does employee treatment and workforce diversity impact corporate innovative efficiency? Available at SSRN
(http://ssrn.com/abstract=2476543)
McGregor, D. (1960). The human side of enterprise. New York, 21, 166.
Meulbroek, L. K., Mitchell, M. L., Mulherin, J. H., Netter, J. M., & Poulsen, A. B. (1990). Shark repellents and managerial myopia: An empirical test. Journal of
Political Economy, 1108–1117.
Miller, K. L., & Monge, P. R. (1986). Participation, satisfaction, and productivity: A meta-analytical review. Academy of Management Journal, 29(4), 727–753.
Oswald, A. J., Proto, E., & Sgroi, D. (2009). Happiness and productivity Discussion paper 4645. Bonn, Germany: Institute for the Study of Labor (IZA).
Pakes, A., & Griliches, Z. (1980). Patents and R&D at the firm level: a first report. Economics Letters, 5(4), 377–381.
Sauermann, H., & Cohen, W. M. (2010). What makes them tick? Employee motives and firm innovation. Management Science, 56(12), 2134–2153.
Scherer, F. M. (1965). Firm size, market structure, opportunity, and the output of patented inventions. The American Economic Review, 1097–1125.
Sevilir, M., & Tian, X. (2012). Acquiring innovation. AFA 2012 Chicago Meetings Paper.