How does the presidential election period affect the performance of the state-owned enterprise in Indonesia?

Abstract

This study investigates the differences between the SOEs and non-SOEs financial performance and how the presidential election affects their performance. This study uses 3,716 firm-year observations for firms listed on the Indonesian Stock Exchange from 2001 to 2014 as the final sample and uses regression to test the hypotheses. In Indonesia, on average, about 25 parties involved in the presidential election in the past three elections. Due to the complexity of the data collection, this study omits the effect of the unique political parties that also could affect the performance of SOE. This study finds that SOEs outperform financial performance of non-SOEs over the sample periods. Interestingly, this study also finds that the excellent financial performance of SOEs disappears around the election period. It indicates that being a board member of state-owned enterprises (SOEs) is a political position rather than a professional position. For policymakers, these results indicate that election periods influence (reduce) the financial performance of SOEs in Indonesia. This study enhances our understanding of how presidential elections affect the performance of SOEs in Indonesia.

Keywords:

• election
• firm performance
• political connections
• state-owned enterprises

JEL classification:

• D72
• G30
• L25

PUBLIC INTEREST STATEMENT

This study investigates the differences between the SOEs and non-SOEs financial performance and how the presidential election affects their performance. This study uses 3,716 firm-year observations for firms listed on the Indonesian Stock Exchange from 2001 to 2014 as the final sample and uses regression to test the hypotheses. In Indonesia, about 25 parties involved in the presidential election in the past three elections. This study finds that SOEs outperform financial performance of non-SOEs over the sample periods. Interestingly, this study also finds that the excellent financial performance of SOEs disappears around the election period. It indicates that being a board member of state-owned enterprises (SOEs) is a political position rather than a professional position. For policymakers, these results indicate that election periods influence the financial performance of SOEs in Indonesia. This study enhances our understanding of how presidential elections affect the performance of SOEs.

1. Introduction

Previous research documents that political connections give useful benefits to firms especially in developing countries (Fisman, 2001; Leuz & Oberholzer-Gee, 2006). However, there are merely a handful of studies that discuss how the connections affect the performance of State-owned Enterprises (SOEs). This paper studies the performances SOEs and examines how the election period affects their performance. Studying the performance of SOEs is important to enhance our understanding of how efficient firms with state ownership are at using the government budget to operate their professional business. We are particularly interested in how the election period in Indonesia will affect their performance.

To investigate this issue, we identify the unique setting of Indonesia where political connections have a strong influence on business. Fisman (2001) found that those connected firms in Indonesia rely upon the power of their connections. This indicates that firms whose political connections are good, they more likely to enjoy the advantages from their connections. Leuz and Oberholzer-Gee (2006) provide evidence that connected firms in Indonesia face difficulties in maintaining domestic financial access provided by their connections when their connections fall from power. Therefore, Indonesia is a good setting for examining the impact of presidential elections on the performance of SOEs. Harymawan and Nowland (2016) find that politically connected firms give response to the changes in political stability and government effectiveness in Indonesia by adjusting their quality of earnings.

The characteristic SOEs in Indonesia are also unique. Based on Law Number 19 of 2003 concerning State-Owned Enterprises, there are three types of state-owned enterprises (SOEs). The first type is a Perusahaan Umum (Perum), which is SOE whose capital is fully owned by the Republic of Indonesia, not divided into shares, and offered for public benefit. The second type is a Perusahaan Perseroan (Persero) is SOE with the form of limited company whose capital consists of at least 51% of shares owned by the Republic of Indonesia and the goal is to obtain profits. The third type is Perusahaan Perseroan Terbuka (Persero Tbk), which is Persero that conducts a public offering in accordance with regulations in the capital market or Persero who sells its shares on the capital market. In this study, the types of SOEs used are Perusahaan Perseroan Terbuka, considering that company-listed reports are widely used in various decisions related to investment and regulation-making.

In the first part of this study, we compare the performance of SOEs to non-SOEs. As the firms with state ownership, it is likely that these types of firms will have more opportunity to receive preferential treatment from the government, i.e.: preferential credit terms; government bailouts; special licenses (export, import, etc.); and government procurement (Blau et al., 2013; Claessens et al., 2008; Faccio et al., 2006; Khwaja & Mian, 2005; Le & Buck, 2011; Yeh et al., 2013). Therefore, we conjecture that SOEs will have significantly better performance than non-SOEs. Using OLS regression analysis, this study finds that SOEs outperform non-SOEs over the sample periods.

Anecdotal evidence indicates that the appointments for Ministry of State-owned Enterprises in Indonesia are more associated with political appointment rather than professional appointment. Following his/her appointment to a ministry, it is also well known that a new Ministry of SOEs in Indonesia will appoint newly connected directors based on their connections to many SOEs following appointment. Under the rent-seeking theory, the appointed directors are expected to support the interests of their connections (ruling party). If this is the case, it is expected that, in some situations (i.e.: election period), there will be an increase of conflict of interest in SOEs since the directors are politically connected. This situation could lead to decreased performance in these SOEs. Hence, in the second part of this study, we hypothesize that the outstanding financial performance of SOEs will vanish around the election period.

This study investigates these issues using 3,716 firm-year observations from the firm listed on the Indonesian Stock Exchange. This study finds that SOEs have significantly higher firm performance than non-SOEs. Consistent with our prediction, this study finds that the excellent financial performance of SOEs disappears around an election period.

This study provides new evidence to emerging literature examining the factors that can affect the decisions and performances of politically connected firms. Leuz and Oberholzer-Gee (2006) suggest that firms which have political connection will access foreign capital when their patronage falls from power. This study focuses on the way the election could affect the performance of politically connected firms.

The sections of this paper are as follows. In section 2, the research hypotheses are elaborated and developed. While in section 3, it elaborates the sample and variables. Section 4 specifies the empirical models and presents the main results and the results of the sensitivity tests. Section 5 summarizes the paper and presents concluding remarks.

2. Literature review and hypotheses development

2.1. Stated-owned enterprises (SOEs) in Indonesia

Based on Law Number 19 of 2003 concerning State-Owned Enterprises explain that SOEs are business entities which all or most of their capital are owned by the state through direct participation from the separated state assets. There are three types of SOEs, the first type is Perusahaan Umum (Perum), which is SOE whose capital is fully owned by the Republic of Indonesia, not divided into shares, and offered for public benefit. The second type is a Perusahaan Perseroan (Persero) is SOE with the form of limited company whose capital consists of at least 51% of shares owned by the Republic of Indonesia and the goal is to acquire profits. The third type is Perusahaan Perseroan Terbuka (Persero Tbk), which is Persero that conducts a public offering in accordance with regulations in the capital market or Persero who sells its shares on the capital market. In practice, ministers are appointed and authorized to represent the government as state shareholders in the Persero and have capital in the Public Corporation. The appointed minister is the Ministry of State-Owned Enterprises. In addition, the minister is also responsible for the appointment and dismissal of Directors and Commissioners, while the requirements and procedures for the appointment and dismissal of members of the Directors and Commissioners are regulated in a Ministerial Decree.

SOEs handle various objectives and roles under the influence of the government in achieving their social and political goals and their financial goals. When market orientation policies are implemented and the role of government is limited, Indonesian SOEs still play an important role in the development of economic and social welfare (Abeng, 2002). This can be seen from their involvement in primary industries, such as defense, banking, transportation and telecommunications, which have strategic implications for the country (Diah, 2003; Siahaan, 2005). In addition, Indonesian SOEs also play an important role in the field of social welfare. The social role of Indonesian SOEs emerged as a consequence of the 1945 Constitution of the Republic of Indonesia Article 33, highlighting the importance of controlling natural resources and production by government authorities (Abeng, 2002; Diah, 2003) and Article 34, underlines the requirements and importance for the state to provide and be responsible for social security systems and public utilities. This shows that the government must remain in the SOE business. In addition, the centralization of SOEs under the Ministry of State-Owned Enterprises aims to improve the performance of SOEs, eliminate bureaucracy and speed up the privatization process needed to resolve the national economic situation (Abeng, 2002).

2.2. Hypotheses development

In the previous studies state that political connections affect the decision-making processes and performances of firms in developing countries (Fisman, 2001; Harymawan & Nowland, 2016; Leuz & Oberholzer-Gee, 2006). Shleifer and Vishny (1994) posit that politicians can influence firms through subsidies. However, as we are aware, only a few papers have discussed firms owned by governments (SOEs), especially in terms of whether a presidential election affects their performance. The unique setting of SOEs provides us with a cleaner measure of political connections. The measure of SOEs as a proxy of political connections is also free from the selection bias problem which also means it I expected to provide more credible results.

Several studies have also investigated how political connections affect the business decision-making. Shirodkar et al. (2018) examine the determinant of political corporate social responsibility (PCSR) of multinational companies (MNC) in India. They found that MNCs are more likely to be engaged in PCSR especially for MNCs’ which subsidiaries are dependent on local resources. Boubakri et al. (2012), in their international study, investigate firm performance and the leverage of politically connected firms. They find that politically connected firms have higher performance and leverage than non-connected ones. Harymawan et al. (2019) find that firms with specific political connections, for example, military connections, in Indonesia, have higher leverage. Yu (2013) investigated the performance of SOEs in China from 2003–2010. Her main finding shows that the performances of SOEs are sensitive to government strategic decisions on SOEs. The split share structure reform in China from 2005–2006 boosted the performance of SOEs. Eforis (2018) and Astami et al. (2010) explain that the level of performance of SOEs with private sector or public ownership is higher than those owned entirely by the government. Interestingly, documents in prior research how that political connections can derive valuable benefits from their connections; we therefore hypothesize that the performance of state-owned enterprises (SOEs) is significantly higher than non-SOEs. The formal hypothesis is as follows:

H1: The financial performance of SOEs is significantly higher than non-SOEs

SOEs’ performance is sensitive to government decisions in SOEs in China (Yu, 2013). Sukmadilaga et al. (2014) also explained that during the Crisis the government would inject direct subsidies to enable SOEs to maintain their performance. Apriliyanti and Kristiansen (2019) show that rent-seeking, collusion and corruption among political and business elites and high-ranking officials in the government detain the reforming of corporate governance of SOEs. Anecdotal evidence shows that the appointments for Ministry of State-owned Enterprises in Indonesia are more political than professional. Several examples are as follows. Firstly, Soeharto appointed Tanri Abeng to the Ministry of SOEs in 1998. Tanry Abeng is known to be one of the key people in the Bakrie Group, a big business group led by the Bakrie family which is closely connected to Soeharto. Another example is the appointment of Lasamana Sukardi to the Ministry of SOEs in 1999. He served in that position in both Wahid’s and Megawati’s presidency. He is known to have been a confidant of Megawati. A recent example is the appointment of Rini Soemarno to the Ministry of SOE. Rini served in this position under Jokowi’s presidential term. Jokowi was supported by Partai Demokrasi Indonesia—Perjuangan (Indonesian Democratic Party of Struggle) for his presidency. Before Jokowi announced his cabinet, the analyst predicted that Rini would be appointed to the Ministry of SOEs because of her close relationship with Megawati, the leader of the Indonesian Democratic Party of Struggle. Based on the above evidence, it is likely that the new Ministry of SOEs will appoint newly connected directors from their connections in many SOEs following their appointment.

The above discussion indicates that the boards of SOEs are highly connected to political positions rather than professional positions. It is expected that in an election period, there will be a high conflict of interest in SOEs which makes it more likely that they will utilise their controlled resources to support the interests of their connections (ruling party). Hence, in the second part of this study, we hypothesise that the financial performance of SOEs will underperform around the election period. The formal hypothesis is as follows:

H2: Excellent financial performance of SOEs will vanish around the election periods

3. Data and method

3.1. Sample and data sources

The initial list of our sample consists of all firms listed on the Indonesian Stock Exchange (IDX) from 2001 to 2014. Then, all financial data was obtained from the OSIRIS database. We constructed time dummy variables to represent each government regime and the election period. We used SAS software 9.3 to handle the data management. Then, we used STATA 13 to conduct regression analyses. We imposed two criteria for the sample selection. Firstly, we required no missing data for all variables used in this study. Secondly, we delete all firms for which the fiscal year did not end in December. Finally, we obtained 3,716 firm-years as our main sample.

3.2. Variable measurements

3.2.1. Firm financial performance

We use two measures for firm financial performance. The first is return on assets (ROA), defined as net income divided by average total assets for the year t. The second is the average of return on assets (AVROA). We construct this measure by calculating the average of return on assets for the lag year (t-1), current year (t), and lead year (t + 1). In addition, we also use return on equity (ROE) and lead of return on assets (LEAD_ROA) as sensitivity checks.

3.2.2. Election period

Next, we construct two dummy variables as election period indicators in Indonesia. The first is ELECTION1. We define this variable as the year of the election period in Indonesia (2004, 2009, and 2014). The second variable is ELECTION2. In this variable, we measure the election period in Indonesia as year t-1 and t (2003, 2004, 2008, 2009, 2013, and 2014).

3.3. Empirical model

This study uses multiple linear regression analysis models with fixed effects to examine the relationship between SOEs and their financial performance and how the presidential election affects their performance. Hypothesis 1 is being tested by specifying the following regression model linking the firm performance variable with state-owned enterprise and batteries of control variables:

(1) $PERFORM={\beta }_1+{\beta }_{2}SO{E}_{jt}+{\beta }_{3}SIZ{E}_{jt}+{\beta }_{4}LEVERAG{E}_{jt}+{\beta }_{5}CAPIN{T}_{jt}+{\beta }_{6}MB{V}_{jt}+INDUSTRYFIXEDEFFECTS+YEARFIXEDEFFECTS+\epsilon$(1)

where PERFORM is one of the financial performance measures. SOE is dummy for state-owned enterprise. Batteries of control variables are based on prior studies (e.g., Bhagat & Bolton, 2008; Core et al., 1999). Some other prior research also find that firm complexity influences the performance of the firm (Andersen & Reeb, 2003; Mashayekhi & Bazaz, 2008). The previous literature also finds that capital structure and intensity are important determinants of firm performance (Andersen & Reeb, 2003; Wu et al., 2012). Core et al. (1999), find that the governance structure will influence the market value of equity and market to book value when the investor values the governance structure. Based on the aforementioned studies, I include firm size (SIZE), firm leverage (LEVERAGE), capital intensity (CAPINT), and market to book value (MBV) as control variables.

To test hypothesis 2, we add the dummy variable indicating the election periods in Indonesia to Equation (1)(1) $PERFORM={\beta }_1+{\beta }_{2}SO{E}_{jt}+{\beta }_{3}SIZ{E}_{jt}+{\beta }_{4}LEVERAG{E}_{jt}+{\beta }_{5}CAPIN{T}_{jt}+{\beta }_{6}MB{V}_{jt}+INDUSTRYFIXEDEFFECTS+YEARFIXEDEFFECTS+\epsilon$(1) . Specifically, the regression model is as follows:

(3) $PERFORM={\beta }_1+{\beta }_{2}SO{E}_{jt}+{\beta }_{3}SOE\ast ELECTIO{N}_{jt}+{\beta }_{4}ELECTIO{N}_{jt}+{\beta }_{5}SIZ{E}_{jt}+{\beta }_{6}LEVERAG{E}_{jt}+{\beta }_{7}CAPIN{T}_{jt}+{\beta }_{8}MB{V}_{jt}+INDUSTRYFIXEDEFFECTS+YEARFIXEDEFFECTS+\epsilon$(3)

where ELECTION is one of the government regime indicators. In this equation, we employ two types of election period indicators: ELECTION1 equal to 1 if a year belongs to the election period (Year 2004, 2009, and 2014), otherwise 0; ELECTION2 1 if a year belongs to the election period and 1 year prior to the election period (Year 2003, 2004, 2008, 2009, 2013, and 2014), otherwise 0.

4. Empirical result and discussion

In this section, we provide descriptive statistics, univariate analyses, and a multivariate analysis for this study. We handle the outlier issues using a winsorizing approach. We winsorize our sample based on the 1 and 99 percentiles of the sample. Following Petersen (2009), we also correct the standard error clustered by firm and year in all regressions for this study.

Panel A in Table 1 presents the distribution of firms by year. As shown in this panel, 144 firms are SOE. The number of SOE firms is increasing over time. Panel B of Table 1 reports the distribution of firms by industry. Nearly 46 percent of SOEs is in the mining industry. The remaining 54 percent is in three industries: Construction; Manufacturing; Transportation, Communications, and Utilities. There are no SOE firms in Agriculture, Forestry, and Fisheries; Wholesale & Retail Trade; Banking and Insurance; Service Industries; and Health, Legal, and Educational Services and the Consulting industry.

Table 1. Firm distribution

Panel A: Firm distribution by year
Year	Non-SOE	SOE		Total
2001	26	0		26
2002	156	6		162
2003	159	7		166
2004	159	8		167
2005	162	8		170
2006	168	8		176
2007	281	11		292
2008	307	11		318
2009	308	11		319
2010	334	13		347
2011	354	14		368
2012	369	15		384
2013	389	16		405
2014	400	16		416
Total	3572	144		3716
Panel B: Firm distribution by industry
Industry		Non-SOE		SOE
		Firms	Percentage	Firms	Percentage
(SIC 0) Agriculture, Forestry, and Fisheries		128	100%	0	0%
(SIC 1) Mining		349	84%	66	16%
(SIC 2) Construction Industries		1024	98%	26	2%
(SIC 3) Manufacturing		641	97%	20	3%
(SIC 4) Transportation, Communications, and Utilities		370	92%	32	8%
(SIC 5) Wholesale & Retail Trade		309	100%	0	0%
(SIC 6) Banking and Insurance		471	100%	0	0%
(SIC 7) Service Industries		254	100%	0	0%
(SIC 8) Health, Legal, and Educational Services and Consulting		26	100%	0	0%
Total		3572	96%	144	4%

This table shows the distribution of SOE and non-SOE firms by year and industry. The sample consists of all firms on the Indonesian Stock Exchange (IDX) listed for the years 2000 to 2014.

The definition of all variables is stated in Appendix A. Table 2 gives the descriptive statistics for the full sample over the sample periods. On average, 4 percent of the samples is SEO firms. The average firm has a return on assets of 5.25 percent, a return on equity of 7.74 percent, leverage of 0.56 percent, capital intensity of 0.35 percent, turnover of 2.21 percent, market to book value of 3.36 percent, and total assets of 439 USD million. The variation of all financial performance variables is relatively high.

Table 2. Descriptive statistics

Variable	Mean	Std. Dev.	Median	Minimum	Maximum
SOE	0.04	0.19	0	0	1
ROA	5.25	12.25	4.52	−36.74	48.34
ROE	7.74	32.63	8.01	−161.30	123.03
AVROA	5.28	10.81	4.58	−27.75	44.39
AVROE	6.94	28.89	8.12	−153.62	92.06
LEAD_ROA	5.47	12.33	4.51	−34.08	50.99
LEAD_ROE	7.29	32.20	7.46	−157.32	123.03
TOTAL ASSETS	439,491	992,081	121,179	389	12,154,823
LEVERAGE	0.56	0.36	0.52	0.02	2.34
CAPINT	0.35	0.24	0.32	0.00	0.90
MBV	3.63	5.49	1.85	0.12	35.89

This table depicts the descriptive statistics for all variables in this study. All firms on the Indonesian Stock Exchange (IDX) listed for the years 2000 to 2014 are the sample of the study.

The results of univariate tests are provided in the next two tables. Pearson correlation matrix is figured in Table 3. The correlation between SOE and ROA is 0.120, and the correlation is highly significant. This suggests that SOE firms have significantly higher financial performance than non-SOEs. The correlation between SOE and other measures of financial performance also shows positive and significant results. As shown in this table, SOE is positively correlated with firm size. We also discover that SOE is negatively related to firm leverage, capital intensity, and market to book ratio. The results of VIF suggest that multicollinearity is not a concern in our regression model.

In Table 4, the full sample (N = 3716) is divided into two subsamples of firm-years for SOE (N = 144) and non-SOE (3,572). As shown in the first row of the table, the return on assets for SOEs is significantly higher than for non-SOEs. The magnitude of SOEs is nearly three times higher than non-SOEs, which is consistent with hypothesis 1. For other measures of firm financial performance, it is also apparent that SOE firms have significantly higher financial performance than non-SOEs. The results of both t-tests and Wilcoxon z-tests for all variables also differ significantly between SOE and non-SOE firms.

Table 3. Pearson correlation matrix

Variable		(1)	(2)	(3)	(4)	(5)	(6)	(8)	(9)	(10)	(11)	(12)
(1)	SOE	1.000	0.120	0.065	0.151	0.087	0.127	0.072	0.216	−0.033	−0.045	−0.078
	p-value		0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.044	0.007	0.000
(2)	ROA	0.120	1.000	0.486	0.884	0.478	0.673	0.324	0.231	−0.312	−0.147	−0.259
	p-value	0.000		0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
(3)	ROE	0.065	0.486	1.000	0.460	0.614	0.364	0.242	0.138	−0.072	−0.098	−0.147
	p-value	0.000	0.000		0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
(4)	AVROA	0.151	0.884	0.460	1.000	0.547	0.845	0.441	0.256	−0.311	−0.147	−0.288
	p-value	0.000	0.000	0.000		0.000	0.000	0.000	0.000	0.000	0.000	0.000
(5)	AVROE	0.087	0.478	0.614	0.547	1.000	0.484	0.614	0.161	−0.127	−0.110	−0.192
	p-value	0.000	0.000	0.000	0.000		0.000	0.000	0.000	0.000	0.000	0.000
(6)	LEAD_ROA	0.127	0.673	0.364	0.845	0.484	1.000	0.513	0.188	−0.207	−0.137	−0.233
	p-value	0.000	0.000	0.000	0.000	0.000		0.000	0.000	0.000	0.000	0.000
(8)	LEAD_ROE	0.072	0.324	0.242	0.441	0.614	0.513	1.000	0.124	−0.068	−0.090	−0.154
	p-value	0.000	0.000	0.000	0.000	0.000	0.000		0.000	0.000	0.000	0.000
(9)	SIZE	0.216	0.231	0.138	0.256	0.161	0.188	0.124	1.000	0.006	0.127	0.026
	p-value	0.000	0.000	0.000	0.000	0.000	0.000	0.000		0.709	0.000	0.109
(10)	LEVERAGE	−0.033	−0.312	−0.072	−0.311	−0.127	−0.207	−0.068	0.006	1.000	0.112	0.393
	p-value	0.044	0.000	0.000	0.000	0.000	0.000	0.000	0.709		0.000	0.000
(11)	CAPINT	−0.045	−0.147	−0.098	−0.147	−0.110	−0.137	−0.090	0.127	0.112	1.000	0.088
	p-value	0.007	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000		0.000
(12)	MBV	−0.078	−0.259	−0.147	−0.288	−0.192	−0.233	−0.154	0.026	0.393	0.088	1.000
	p-value	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.109	0.000	0.000

This table shows the Pearson correlation for all variables in the study. The sample encompasses all firms on the Indonesian Stock Exchange (IDX) listed for the years 2000 to 2014.

Table 4. Firm characteristics

Variable	Non-SOE	SOE	Mean test	Non-SOE	SOE	Wilcoxon
	mean	mean	t-statistics	median	Median	z- statistics
ROA	4.954	12.549	7.73***	4.245	8.065	8.04***
ROE	7.316	18.367	7.7***	7.500	18.250	7.17***
AVROA	4.944	13.334	8.77***	4.240	8.583	8.70***
AVROE	6.436	19.470	10.13***	7.627	20.087	8.15***
LEAD_ROA	5.155	13.219	7.51***	4.195	9.185	8.00***
LEAD_ROE	6.829	18.703	7.8***	7.140	17.900	7.20***
SIZE	11.587	13.484	18.47***	11.602	13.590	12.93***
LEVERAGE	0.563	0.501	−3.53***	0.524	0.512	−1.12
CAPINT	0.357	0.301	−2.91***	0.326	0.259	−2.62***
MBV	3.711	1.499	−13.43***	1.926	1.047	−7.01***

This table shows the firm characteristics of SOE relative to non-SOE firms in the cross-sectional sample in this study. The sample comprises IDX-listed firms from 2000 to 2014 that have financial data available from OSIRIS. Results of mean t-tests and Wilcoxon (z-tests) are displayed. Significance at ***1 percent.

From this section onwards, we discuss the results of multivariate analyses in this study. Table 5 presents the results of regression for firm financial performance on SOE. Two specifications are provided in this table. In the first specification, we use ROA as the firm financial performance measure. It comes up with the coefficient of ROA is 2.768 and it is significant at the level of 1 percent (t = 2.88). In the second specification, we use AVROA as the firm financial performance measure. Similarly, to the first specification, we find the coefficient of AVROA is 3.887 and significant at the 1 percent level (t = 4.19). Consistent with our prediction, the SEO is positive and significantly related to firm financial performance. These findings indicate that SOEs have significantly higher performance than non-SEOs. Of the control variables, we find that firms with higher performance are associated with larger firms, lower firm leverage, lower capital intensity, and lower market to book value. It comes up with the control variables are consistent with the previous literature.

Table 5. State-owned enterprise and firm performance

Variable	Pred.	ROA	AVROA
	Sign	(1)	(2)
SOE	+	2.768***	3.887***
		(2.88)	(4.19)
SIZE	+	1.894***	1.828***
		(14.02)	(15.29)
LEVERAGE	-	−8.031***	−6.910***
		(−8.76)	(−9.97)
CAPINT	-	−8.811***	−7.761***
		(−10.36)	(−9.87)
MBV	-	−0.356***	−0.355***
		(−8.32)	(−9.63)
INTERCEPT	?	−7.636***	−6.375**
		(−2.69)	(−2.15)
Industry dummies		Included	Included
Year dummies		Included	Included
Adjusted R2		0.215	0.244
N		3716	3241

Regression models relating to firm performance, state-owned enterprise, and firm-specific control variables in this study. The dependent variable is one of the performance measures. The sample encompasses all IDX-listed firms from 2000 to 2014 that have financial data available from OSIRIS. Standard errors are clustered by firm and year. t-statistics are in parentheses. All continuous variables are winsorized at the 1 percent and 99 percent levels. Significance at **5 percent, ***1 percent.

SOEs' capital is obtained through direct investments that originate from state assets. It indicates that SOEs will have more opportunity to receive preferential treatment from the government (Blau et al., 2013; Claessens et al., 2008). Therefore, SOEs have significantly higher performance than non-SEOs. This result consistent with Yu (2013) who explains that the performances of SOEs are sensitive to government strategic decisions on SOEs. The government would inject direct subsidies to enable SOEs to maintain their performance (Sukmadilaga et al., 2014). Eforis (2018) and Astami et al. (2010) also show that SOEs that are partly owned by the public demonstrate higher performance levels when compared to SOEs entirely owned by the government. Thus, it can be concluded that SOEs have the power of political connections to affect the business decision-making.

Table 6 displays the results of regression for firm performance on SOE around the election periods in Indonesia. If the election periods affect the financial performance of SOE, it is expected to discover a difference between the interaction of SOE and the election period indicator and the SOE variable. We employ two types of election indicators in this table, ELECTION1 and ELECTION2. In the first specification, we estimate Equation (3)(3) $PERFORM={\beta }_1+{\beta }_{2}SO{E}_{jt}+{\beta }_{3}SOE\ast ELECTIO{N}_{jt}+{\beta }_{4}ELECTIO{N}_{jt}+{\beta }_{5}SIZ{E}_{jt}+{\beta }_{6}LEVERAG{E}_{jt}+{\beta }_{7}CAPIN{T}_{jt}+{\beta }_{8}MB{V}_{jt}+INDUSTRYFIXEDEFFECTS+YEARFIXEDEFFECTS+\epsilon$(3) using the first measure of the election period, ELECTION1. The coefficient of SOE is 2.992 and significant at the 1 percent level (t = 2.69). Interestingly, it is found that the coefficient of interaction SOE*ELECTION1 is negative and insignificant. In the second specification, we re-estimate Equation (3)(3) $PERFORM={\beta }_1+{\beta }_{2}SO{E}_{jt}+{\beta }_{3}SOE\ast ELECTIO{N}_{jt}+{\beta }_{4}ELECTIO{N}_{jt}+{\beta }_{5}SIZ{E}_{jt}+{\beta }_{6}LEVERAG{E}_{jt}+{\beta }_{7}CAPIN{T}_{jt}+{\beta }_{8}MB{V}_{jt}+INDUSTRYFIXEDEFFECTS+YEARFIXEDEFFECTS+\epsilon$(3) using the second measure of the election period, ELECTION2. The coefficient of SOE is 3.636 and is significant at the 1 percent level (t = 2.86). Similarly, to the result in the first specification, we find the coefficient of interaction SOE*ELECTION1 is negative and insignificant. These findings suggest that the excellent performance of SOE vanishes during the election period. This indicates that the election periods in Indonesia affect the financial performance of Indonesian SEOs. The results for AVROA in the last two specifications are similar to the results for ROA.

Table 6. State-owned enterprise, politically connected firms, firm performance, and presidential election

Variable	Pred.	Dependent variable: ROA		Dependent variable: AVROA
	sign	(1)	(2)	(3)	(4)
SOE	+	2.992***	3.636***	3.756***	4.201***
		(2.69)	(2.86)	(3.75)	(3.58)
SOE*ELECTION1	-	−0.922		0.879
		(−0.46)		(0.38)
SOE*ELECTION2	-		−1.809		−0.758
			(−1.00)		(−0.43)
ELECTION1	?	1.561		2.074
		(0.72)		(0.85)
ELECTION2	?		1.571		1.398
			(0.72)		(0.58)
SIZE	+	1.894***	1.894***	1.829***	1.828***
		(14.02)	(14.02)	(15.29)	(15.29)
LEVERAGE	-	−8.030***	−8.028***	−6.909***	−6.909***
		(−8.76)	(−8.76)	(−9.97)	(−9.97)
CAPINT	-	−8.811***	−8.809***	−7.761***	−7.760***
		(−10.36)	(−10.36)	(−9.86)	(−9.86)
MBV	-	−0.356***	−0.356***	−0.355***	−0.355***
		(−8.32)	(−8.32)	(−9.63)	(−9.63)
INTERCEPT	?	−7.637***	−7.640***	−6.378**	−6.374**
		(−2.69)	(−2.69)	(−2.15)	(−2.15)
Industry dummies		Included	Included	Included	Included
Year dummies		Included	Included	Included	Included
Adjusted R2		0.214	0.215	0.243	0.243
N		3716	3716	3241	3241

Regression models relate the firm performance, state-owned enterprise, election period, and firm-specific control variables in this study. The dependent variable is one of the performance measures. The sample comprises all IDX-listed firms from 2000 to 2014 that have financial data available from OSIRIS. Standard errors are clustered by firm and year. t-statistics are in parentheses. All continuous variables are winsorized at the 1 percent and 99 percent levels. Significance at **5 percent, ***1 percent.

SOE performance is sensitive to political environment and government decisions (Yu, 2013). During the election period, political environment in Indonesia will fluctuate because there is competition between political parties. There will be a high conflict of interest within SOEs which makes it more likely that they will use their controlled resources to support the interests of their connections. Furthermore, it is possible that the new SOE Ministry will appoint new directors who are connected from their connections in many SOEs after their appointment. Therefore, SOE's excellent performance will disappear during the election period. The results of this study are in line with Apriliyanti and Kristiansen (2019) which shows that rent-seeking among political and business elites in the government will hamper the governance and performance of SOE companies. In addition, SOEs are companies that have political connections with the government, so their performance will depend on the strength of their connections (Fisman, 2001; Harymawan & Nowland, 2016; Leuz & Oberholzer-Gee, 2006).

5. Conclusion

This study aims to investigate the differences between the SOEs and non-SOEs financial performance and whether the performance of SOEs, major players in Indonesia’s economy, is affected by presidential elections for firms listed on Indonesian Stock Exchange from 2001–2014. The result shows that SOEs have better performance of financial than non-SOE firms during the sample period because of several factors such as government procurement advantage and higher financial leverage. Furthermore, this study finds there is no difference in performance between SOEs and non-SOE firms along the election period. It indicates that the excellent performance of SOE vanishes during the election period because of a high conflict of interest within SOEs during the election periods. This study shows new evidence of the dynamic impact of the presidential election on the performance of SOEs in Indonesia.

This paper contributes to the literature on political connections by investigating the financial performance of SOEs in Indonesia which has a high level of corruption and lobbying, especially during elections. This study also strengthens our understanding of institutional factors that facilitate the existence and benefits of political connections in emerging markets. Future studies must continue to examine other factors that cause the differences in the performance of SOEs and non-SOEs, such as changes in government policies and market reforms. This can be done in other single markets from time to time or in cross-country settings.

Additional information

Funding

This research was supported by Universitas Airlangga under 2017 Faculty of Economics and Business Research Grant.

Notes on contributors

Iman Harymawan

Iman Harymawan is an Assistant Professor in Accounting Department of Universitas Airlangga. He obtained his Ph.D. from the City University of Hong Kong and MBA from National Cheng Kung University. Research interest: political and military connections in business.

Mohammad Nasih

Mohammad Nasih is the Professor of Accounting Department of Universitas Airlangga. He obtained his Doctoral degree from Universitas Airlangga and Master degree from Institut Teknologi Bandung in indonesia. Research interest: financial accounting.

Novrys Suhardianto

Novrys Suhardianto is the Lecturer of Accounting Department of Universitas Airlangga. He obtained his Doctoral degree from City University of Hong Kong and Master degree from Universitas Brawijaya Indonesia. Research interest: financial accounting.

Elvia Shauki

Elvia Shauki is the Lecturer of Financial Accounting, School of Commerce, University of South Australia. She obtained her Doctoral degree from La Trobe University Australia and Master degree from Catholic University of Louvain in Belgium. Research interest: financial accounting and integrated reporting.

Appendix A.

Variable Description

Table

Variable		Definition	Data Source
Dependent variable:
ROA	=	net income divided by average total assets for the year t;	OSIRIS
LEAD_ROA	=	the return on assets of firm j in year t + 1;	OSIRIS
AVROA	=	the average of return on assets of firm j in year t-1, t and t + 1;	OSIRIS
ROE	=	Net income after preferred divided by average total equity for the year t;	OSIRIS
LEAD_ROE	=	the return on equity of firm j in year t + 1;	OSIRIS
AVROE	=	the average of return on equity of firm j in year t-1, t and t + 1;	OSIRIS
Test variables:
SOE	=	1 is the firm is state-owned enterprise and otherwise 0;	-
ELECTION 1	=	1 if a year belong to the election period (Year 2004, 2009, and 2014), otherwise 0	-
ELECTION 2	=	1 if a year belong to the election period and 1 year prior to election period (Year 2003, 2004, 2008, 2009, 2013, and 2014), otherwise 0	-
Control variables:
SIZE	=	natural logarithm of total assets;	OSIRIS
LEVERAGE	=	total liabilities scaled by total assets;	OSIRIS
CAPINT	=	total property, plant, and equipment divided by total assets;	OSIRIS
MBV	=	market to book value;	OSIRIS