The Effect of Board Diversity on Earnings Management: Evidence
from Property, Real Estate, and Construction Firms in Indonesia
Rifdah Yudi and Iman Harymawan
Department of Accountancy, Universitas Airlangga, Indonesia
Keywords: Board Diversity, Gender, Nationality, Education, Earnings Management.
Abstract: The aim of this study is to examine the effect of board diversity on earnings management using evidence of
property, real estate, and construction firms listed in Indonesia Stock Exchange in the period of 2010 to 2015.
Multiple linear regression is employed in the research as analysis technique while data is processed using
STATA/MP 14.2 software. The result shows that diversity in nationality and education can reduce the level
of earnings management in a company. However, the gender diversity does not significant negative effect on
earnings management, while the board nationality diversity and education diversity proven to have significant
negative effect on earnings management.
1 INTRODUCTION
Economic development in Indonesia is growing
rapidly and competition between individuals or
companies are also increasingly stringent, requiring
individuals to be more efficient and effective in
managing the company and to produce competent
human resource. As one country that adopt ASEAN
Economic Community agreement, Indonesia is
expected to anticipate the consequences of free-flow
of goods, services, and labor. Opening the doors for
workers to have easier access which allow transfers
of labor in terms of physical and psychological state,
enabling inclusive environment for diverse Board of
Directors.
Board Diversity can be defined as the degree of
diverse traits which possessed by individuals who
made up the council. Moreover, diversity can be
defined as diverse elements that construct the
structure and composition of boards including age,
gender, nationality, cultural background, educational
background, technical skills, and expertise, life
experiences, personal attitudes, and perspectives
(Sener and Karaye, 2014). Diverse traits in the
composition of board of director helps companies
identify and fix issues within the firms, provide
diverse human resources with abundant insights,
enhance confidence of potential investors, and
improving corporate performance.
Board of commissioners and directors of the
company is responsible for monitoring the quality of
the information contained within the financial
statements. Despite having the obligation of
providing assurance of truth and fair representation of
financial statements, managers may have incentive to
conduct earnings management which may obscure
investors’ rational decision about the real
performance of the firm. Diverse council could, to
some extent, mitigate the risk of opportunistic
management to engage in earnings management
practices. Peni and Vahama (2010) and Garvious, et
al. (2012) stated that the existence of one of the
female CEO or CFO women will reduce the level of
earnings management. Research conducted by Sander
van den Berg (2015) in his research to develop a
hypothesis that nationality significant negative effect
on earnings management. But in the end, the study get
the result that national diversity has a positive effect
on earnings management which means that the
previous hypothesis is rejected. Research conducted
by Rashidah and Fairuzana (2006) which shows that
the competence of the audit committee members has
a negative correlation with earnings management
(discretionary accrual).
Understanding the pervasive impact of earnings
management upon investors’ ability to discern the
real condition of the financial statement, this research
revolve around the following problem:
306
Yudi, R. and Harymawan, I.
The Effect of Board Diversity on Earnings Management: Evidence from Property, Real Estate, and Construction Firms in Indonesia.
In Proceedings of the 1st International Conference on Islamic Economics, Business, and Philanthropy (ICIEBP 2017) - Transforming Islamic Economy and Societies, pages 306-311
ISBN: 978-989-758-315-5
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

1. Does board gender diversity affect the level of
earnings management?
2. Does nationality of board diversity affect the
level of earnings management?
3. Does education board diversity affect the level of
earnings management?
2 THEORETICAL REVIEW
2.1 Theory of Agency
According to Jensen and Meckling (1976), agency
theory describes a relationship arising from a contract
between the principals and others who called the
agency, where the principals delegate a job to the
agent.
Eisenhardt (1989) argues that there are two
problems, namely; different desire or purpose
between the agent and the principal as well as
asymmetric information of business management
perceived by principals).
Agency theory proposed by Jensen and Meckling
(1976) further explained that if the two parties, both
the agent and the principal are utility maximizers, and
then the agent does not necessarily act in accordance
with the interest of the principal. Agents are
motivated to maximize their interest for example, by
retaining their managerial position, remuneration,
and other managerial position benefits. This is
contrary to the interests of the principals who seek to
maximize the return on its resources, so as to give rise
to a conflict of interest between the agent and the
principal.
3 HYPOTHESES DEVELOPMENT
3.1 Influence of Board Gender
Diversity to Earning Management
The emergence of women as both a leader in business
and society, and occupies a major portion than men,
The importance of the representation of women
on boards of companies and their role in providing
and securing resources have recently claimed to be
clear (Burke and Mattis, 2000). According to research
by Garvious et al. (2012) suggest that women tend to
be more ethical to make judgments and to behave than
men, and the presence of women in top management
will provide a drag on earnings manipulation
practices because of the involvement of gender issues
in it. Therefore, researchers developed the hypothesis
as follows:
H1: Board gender diversity negatively affect
earnings management.
3.2 Influence of Board Nation
Diversity to Earning Management
Boards with different nationalities introduce
heterogeneity of ideas, experiences, and perspectives.
Furthermore, diversity in board can reduce the
asymmetry of information and agency costs
associated; increase domestic company's financial
flexibility by increasing the potential investors and
financing opportunities; and expand the cross-border
flow of knowledge and technology (Fogel et al.,
2013). Sander van den Berg (2015) in his research to
develop a hypothesis that nationality significant
negative effect on earnings management. But in the
end, the study get the result that national diversity has
a positive effect on earnings management which
means that the previous hypothesis is rejected.
Therefore, researchers developed the hypothesis as
follows:
H2: Nation board diversity negatively affect
earnings management.
3.3 Influence of Board Education
Diversity to Earning Management
Education is high on the board, to be positively
associated with attitudes toward new products and
innovations. A board with higher secondary
education may be more adaptive and competent to
develop new strategies to cope with continuously
changing market conditions in developing
countries. The results of Rashidah and Fairuzana
(2006) which shows that the competence of the audit
committee members has a negative correlation with
earnings management (dicretionary accrual).
Therefore, researchers developed the hypothesis as
follows:
H3: Board diversity education negatively affect
earnings management.
4 DATA AND METHODS
Data used in this study is panel data which is a
combination of time series and cross section and
obtained from the company’s audited annual report.
Data of companies listed in Indonesia Stock
Exchange (IDX) during the period of 2010 to 2015.
The Effect of Board Diversity on Earnings Management: Evidence from Property, Real Estate, and Construction Firms in Indonesia
307

This study uses a quantitative approach as a
method of research that is based on the philosophy of
positivism, is used to examine the population or a
particular sample, sampling technique is generally
done at random, data collection using the instrument
of research, analysis of quantitative data / statistics,
with the aim to test the hypothesis that has been set
(Sugiyono, 2014). Quantitative research was
conducted with a level of associative explanation by
using the approach of explaining the positions of
variables that are studied and the relationship between
the variables with other variables through hypothesis.
4.1 Data analysis technique
Before performing data processing of descriptive
statistics, each variable will firstly be winsorizing to
cope with the data having outliers. Winsorize is
chosen to pull out outliers of data set for the level of
1% of the lowest and the highest, leaving dataset rate
of 99%.
4.2 Descriptive Statistics Analysis
Descriptive statistics were used to describe and
provide a general overview of statistical research data
for each variable in the study. Measures of central
tendency of data to be used in this research is the
average value (mean), standard deviation, minimum
value and a maximum value.
4.3 Multiple Linear Regression Analysis
The analysis used in this research is multiple linear
regression analysis. This method is an analytical tool
forecasting the effect of two or more independent
variables on a dependent variable in which to prove
the presence or absence of a functional relationship or
causal relationship between two or more variables
with one dependent variable (Riduwan, 2010). In this
case the independent variable is board diversity
represented by gender, nationality, and education.
While the dependent variable is earnings
management. Multiple linear regression equation in
this study can be formulated as follows:
DA = α + β
1
GEN + β
2
NAT + β
3
EDU+ β
4
FSIZE
+ β
5
ROA + β
6
LEV + ε
For this regression, the study used two regression
models, OLS and OLS with robust. When using
multiple linear regression analysis there is still a
possibility heteroscedasticity symptoms. Therefore,
the researchers used multiple linear regression
analysis with robust to address the problem. The
method is robust regression method is used when
there is a distribution of the errors are not normal or
there are some outliers that will affect the model
(Ryan, 1997). This method is an important tool to
analyze data that is affected because of the outliers so
that the resulting models are robust or resistance
against outliers.
5 RESULTS
5.1 Results Descriptive Statistics
Table 1: Results Descriptive Statistics.
Variable
Mean
Median
Std. Dev.
Min
Max
DA
0.088
0.061
0.085
0.001
0.419
GEN
0.165
0.143
0.152
0.000
0.600
NAT
0.043
0.000
0.069
0.000
0.286
EDU
0.372
0.333
0.216
0.000
0.917
FSIZE
28.056
28.541
2.161
21.765
31.215
ROA
0.056
0.047
0.057
-0.092
0.254
LEV
0.419
0.448
0.189
0.008
0.850
Source: Processed Data Output STATA.
a. The level of earnings management of data
distribution (discretionary accrual) has a degree
of variation of 96, 59%. This shows earnings
management (discretionary accrual) company
overall sample is uniform, in which the value of
earnings management (discretionary accrual) of
each sample company is relatively close to the
average of earnings management (discretionary
accrual).
b. Data distribution board level gender have this
level of variation of 92, 12%. This demonstrates
the gender proportion of board members in the
ICIEBP 2017 - 1st International Conference on Islamic Economics, Business and Philanthropy
308

company's overall sample is uniform, where the
value of the proportion of female board
members in the company by other companies
relatively similar or homogenous.
c. Data distribution board level nationality have
this level of variation of 160, 47%. It shows the
proportion of nationality for board members in
the company as a whole is different samples, in
which the difference in the proportion of board
members who are foreign nationals overall is
very volatile and there is a difference in large
numbers between one company with another
company.
d. Data distribution board level of education has a
degree of variation of 58, 06%. It shows the
proportion of board members based on the level
of education in the company's overall sample is
uniform, where the value of the proportion of
board members who have the education level of
the Master level in one company with another
company relatively similar or homogenous.
e. The level distribution of the data firm size
(FSIZE) has a degree of variation by 7, 70%.
This shows the firm size (FSIZE) the company
overall sample is uniform, in which the value of
firm size (FSIZE) one company with another
company close to the average firm size (FSIZE).
f. Data distribution level return on
assets (ROA) has a degree of variation
of 101, 79%. It shows the return on
assets (ROA) the company overall sample is
different, where the value of return on
assets (ROA) each sample company in a
predetermined observation period showed a
fluctuating value. This proves the company's
ability to manage the assets to generate profits
for each firm is relatively different samples.
g. Data distribution level of leverage (LEV) has a
degree of variation of 45, 10%. This
demonstrates the leverage (LEV) the company
overall sample is uniform, where the level
of leverage of the company with other
companies in a predetermined observation
period is relatively the same. This indicates that
each sample company policy
is the same relative to the use of debt to assets of
the company.
5.2 Results of Multiple Linear
Regression Analysis Model
Table 2: Results of Multiple Linear Regression Analysis.
Variables
Predictio
n
Relation
ship
DA
Multiple
Linear
Regression
Robust
regression
GEN
-
-0.029
(-0.92)
-0.029
(-1.08)
NAT
-
-0.129 *
(-1.81)
-0.129 *
(-1.95)
Communities
-
-0.056 **
(-2.32)
-0.056 **
(-2.55)
FSIZE
+
0,001
(0.53)
0,001
(0.64)
ROA
+
0,302 ***
(3,62)
0,302 ***
(2.60)
LEV
-
-0.001
(-0.03)
-0.000
(-0.03)
year dummies
Including
Including
adjusted R2
0.241
R-Squared
0.272
0.272
N
265
265
Source: Data Output STATA Sports
Big contribution of variable gender, nationality,
education, firm size (FSIZE), return on
assets (ROA) and leverage (LEV) to earnings
management amounted to 27.23%. While the rest of
72.77% is explained by other variables that are not
included in this study.
a. Gender board regression coefficient of -0.029
and have a negative effect on earnings
management. H al indicate if gender-
board increase one time, then the earnings
management will be decreased by 0,029, and
vice versa, assuming other variables held
constant. This variable has a p value of 0.280
indicates that it is not significant.
b. Nationality board regression coefficient of -
0.129 and have a negative effect on earnings
management. H al indicate if
the board increases
one nationality, then earnings management will
be decreased by 0.129, and vice versa, assuming
other variables held constant. This variable has
a p value of 0.052 indicates that significant at
the 10% level.
c. Regression coefficient of -0.056 and the board
of education has a negative effect to earnings
management. H al indicate if the board of
education increased by one, then earnings
The Effect of Board Diversity on Earnings Management: Evidence from Property, Real Estate, and Construction Firms in Indonesia
309

management will be decreased by 0.056, and
vice versa, assuming other variables held
constant. This variable has a p value of 0.011
indicates that significant at the 5% level.
d. The regression coefficient firm size (FSIZE) of
0.001 and has a positive influence on earnings
management. H al indicates if the firm
size (FSIZE) increased by one, then earnings
management will increase by 0,001, and vice
versa, assuming other variables held
constant. This variable has a p value of 0.523
indicates that it is not significant.
e. Regression coefficient value return on
assets (ROA) of 0,302 and has a positive
influence on earnings management. H al
indicate if the return on assets (ROA) increased
by one, then earnings management increased by
0,302, and vice versa, assuming other variables
held constant. This variable has a p value of
0.010 indicates that significant at the 1% level.
f. The regression coefficient of leverage (LEV)
amounted to - 0.001 and has a negative effect on
earnings management. H al indicate
if leverage (LEV) increased by one, then
earnings management decreased by 0,001, and
vice versa, assuming other variables held
constant. This variable has a p value of 0.976
indicates that it is not significant.
6 DISCUSSION
6.1 Influence of Board Gender
Diversity to Earning Management
The results showed that board’s gender diversity has
negative and insignificant effect to earnings
management. This is because the directors with
female gender would avoid the risk caused by these
opportunistic actions and agency problems between
agents and principals. This has resulted in the
directors with female gender to avoid aggressive
earnings management practices in accounting
policies.
However, companies having male director seeks
to avoid the practices of earnings management. This
is because the board of directors with the male gender
more have the ability to run operations thoroughly, so
that practices such as the manipulation of earnings
management can be minimized. Due to the similarity
of standpoint, gender alone is not enough to predict
the actions of management. According to research
conducted by Peni and Vahama (2016), it can be
inferred that the presence of female CEO or CFO
women will reduce the level of earnings management.
6.2 Influence of Board Nationality
Diversity to Earning Management
The results shows that nationality and diversity of
board of directors have insignificant and negative
effect on the earnings management. The result
indicate that diverse nationality of board of directors
may reduce earnings management. Furthermore,
heterogeneous group provides diverse ideas and
knowledge from wide range of perspective leading
towards more constructive discussion which is useful
for board’s dynamic within the firms. The
composition of the board of directors which entails
different backgrounds of individuals can improve and
maintain management’s credibility, mitigating the
incentives to engage in earnings management
practices.
Sander van den Berg (2015) in his research
develop a hypothesis that diverse nationality has
significant negative effect on earnings management
yet, the study finds that nationality has positive effect
on earnings management which means that the
previous hypothesis is rejected.
6.3 Influence of Board Education
Diversity to Earning Management
The board diverse education background proven to
have significant and negative effect on earnings
management which translate into the reduction of
earnings management. The more members of the
board who have higher secondary education, the
lower the level of earnings management. So the
diversity of educational backgrounds and experiences
are important to the composition of the board as a
whole, because ultimately affect earnings
management in the company. The results of the study
concurs with research conducted by Rashidah and
Fairuzana (2006) shows that the competence of the
audit committee members has a negative correlation
with earnings management (dicreationary accrual).
REFERENCES
Burke, R. J. and M. C. Mattis: 2000, ‘Women Corporate
Boards of Directors: Where Do We Go from Here?’, in
R. J. Burke and M. C. Mattis (eds.), Women on
Corporate Boards of Directors (Kluwer Academic
Publishers, The Netherlands), pp. 3–10
ICIEBP 2017 - 1st International Conference on Islamic Economics, Business and Philanthropy
310

Eisenhardt, Kathleen M. 1989. Agency Theory: An
Assesment and Review. The Academy of Management
Review, Vol. 14, No. 1. 1989.
Fogel, K.S., Lee, K.K., Lee, W.Y. and Palmberg, J. 2013.
Foreign Direct Investors as Change Agents: The
Swedish Firm Experience. Corporate Governance: An
International Review. 21(6): 516-534.
Garvious, Ilanit Einav Segev Rami Yosef. 2012. Female
Directors and Earnings Management In High-
Technology Firms. Pacific Accounting Review, Vol. 24
Iss 1 pp. 4-32.
Jensen, Michael C., William H Meckling. 1976. Theory of
the Firm: Managerial Behavior, Agency Costs and
Ownership Structure. Journal of Financial Economics
(JFE), Vol. 3, No. 4, 1976.
Peni, Emilia and Sami Vahamaa. 2010. “Female Executive
and Earnings Management”. Managerial Finance
Journal, Vol.36, No.7, pp. 629-645.
Rashidah Abdul Rahman Fairuzana haneem Mohamed Ali.
2006. Board, Audit Committee, Culture and Earnings
Management: Malaysian Evidence. Managerial
Auditing Journal, Vol. 21 Iss 7 pp.783-804.
Riduwan. 2010. Skala Pengukuran Variabel-Variabel
Penelitian. Alfabeta, Bandung.
Ryan, T. P. 1997. Modern Regression Methods. Canada :
John Wiley & Sons,Inc
Sander van den Berg. 2015. “Nationality Diversity of The
Board of Directors and The Level of Earnings
Management”.
Sener, Irge and Abubakar B. Karaye. 2014. Board
Composition and Gender Diversity: Comparison of
Turkish and Nigerian Listed Companies. Social and
Behavioral Sciences 150, 1002 – 1011.
Sugiyono. 2014. Metode Penelitian Bisnis. Alfabeta.
Bandung.
The Effect of Board Diversity on Earnings Management: Evidence from Property, Real Estate, and Construction Firms in Indonesia
311