The Influence of Public Welfare and Audit Findings on Audit Opinion: Empirical Evidence from Provincial Data in Indonesia

Abstract

The aim of the study is to empirically investigate and analyze the influence of public welfare, audit findings, and follow-up of audit recommendations on audit opinion with the disclosure level of financial reports as an intervening variable using agency theory and signaling theory. To achieve this purpose, a quantitative research method was employed. Population of this study is Provincial Government Financial Reports in Indonesia for fiscal years 2016 to 2018. There were 84 financial reports that met the criteria of purposive sampling. The data were gathered from the websites of the Audit Board of the Republic of Indonesia and the Indonesian Central Bureau of Statistics. In this study, the hypothesis-testing tool is path analysis using the Statistical Package for Social Sciences version 15. Based on the multiple regression analysis, the results show that audit findings, public income, and the disclosure level of financial reports significantly influenced audit opinion. Besides, the follow-up of audit recommendations and public health significantly influenced audit opinion through the disclosure level of financial reports. This study suggests that, in order to have better audit opinion, local governments need to improve public welfare, follow-up audit finding, and disclose more details in financial report.

1. Introduction

The implementation of good public governance in local governments in Indonesia is increasingly needed since the 1999 political reform after Soeharto stepped down. Bureaucratic reform and the decentralization of authority caused local governments to gain enormous rights in organizing and managing the resources. Thus, they need to improve the implementation of good public governance in local government (Ahyaruddin & Akbar, 2018). Act number 17 in the year 2003 of State Finance of the Republic of Indonesia has ordered the governor/regent/mayor as the head of the local government to provide and submit financial reports of their local government. As a reporting entity, the local government is required to carry out accounting and to prepare financial reports comprising Budget Realization Report, Financial Position, Cash Flow Statements, and Notes to Financial Reports prepared according to Government Accounting Standards (Firdaus et al., 2015). These financial statements were then subject to the audit process undertaken by the Audit Board of the Republic of Indonesia (Kusumawati & Ratmono, 2017).

The Audit Board provides an opinion on financial reports based on the results of the examination, which contains audit findings, conclusions, and recommendations. Each finding comprises identified problems in terms of weaknesses in the internal control system (ICS) and/or non-compliance with legislation. Financial reports audited by the Audit Board are the implementation of the accountability principle that prevents misappropriation of authority (BPK, 2015; Yaya & Suprobo, 2019). For the 2016 examination on 537 Local Government Financial Reports (LGFR), the Audit Board revealed 375 (70%) LGFR as unqualified opinions, 139 (26%) LGFR as qualified opinions, and 23 (4%) LGFR as disclaimer opinions. When compared with the achievements in 2016, the quality of LGFR in 2017 has increased with an increase in an unqualified opinion by 6 percentage points, from 70% in LGFR in 2016 to 76% in LGFR in 2017. The quality of LGFR in 2018 has improved as indicated by an increase in an unqualified opinion by 6 percentage points, from 76% in LGFR in 2017 to 82% in LGFR in 2018 (BPK, 2019).

Audit opinion is based on its conformity to Government Accounting Standards (GAS), adequacy of disclosure, compliance with laws and regulations, and effectiveness of ICS. The good quality of financial reports is reflected in the higher level of financial reports disclosure. The probability of unqualified audit opinion is higher if adequacy disclosure level of financial reports is high (Sari et al., 2015). Based on the explanation above, this research investigates three main themes that influence provincial government audit opinions, namely, public welfare, audit findings, and follow-up of audit recommendations by incorporating disclosure level as intervening variable. The use of disclosure level as intervening variable is still under-researched. This model of research is expected to contribute to better understanding of the extent to which public welfare and other relevant factors affect audit opinion.

2. Theoretical Framework

2.1. Agency Theory

Agency theory explains the connections between the principals and the agents that have been engaged in making decisions as well as acting on their behalf (Mahaney & Lederer, 2003). In the government sector, agency problem occurs between government officials and the voters as the agents. In this sector, the government is appointed and chosen to be the principal. Government, as the party to public services, has more important information to be used to make policies and decisions. The policies will emphasize and prioritize the government and authority only and ignores the public’s interest and welfare. In order to minimize the problem, presenting transparent and accountable financial reports should be done by the local government (Adiputra et al., 2018). The form of accountability for the implementation of government through financial reports to the public is the agency’s relationship between government as agent and society as principal. Financial reports through the Internet are an attempt to reduce information asymmetry between government as agent and society as principal (Laswad et al., 2005; Sutaryo & Sinaga, 2018).

2.2. Signaling Theory

With signaling theory, quality entities will signal users of the report by providing complete disclosure or information (Priharjanto & Wardani, 2016). Evans and Patton (1987) in Arifin and Fitriasari (2014) stated that, in the context of government, the government is trying to give a good signal to the public so that the public can continue to support government activities that are currently running. The government will give signals to the public by providing quality financial reports, improving internal control systems, more complete disclosures, and more detailed explanations in disclosures. The government can also package more information on achievement and financial performance in order to demonstrate that the government has carried out the mandate given by the public (Hendriyani & Tahar, 2015).

2.3. Audit Opinion

Article 1 of the Act Number 15 of 2004 explains the opinion is a professional statement based on the conclusion of the examiner on the degree of fairness in the information that were included in financial reports. The fairness concerns materiality, financial position and cash flow. The audited financial reports will be used by financial reports’ users in making decisions or policies to increase the quality of financial reports itself. The types of audit opinion by the Audit Board are Unqualified Opinion including Modified Unqualified Opinion, Qualified Opinion, Adverse Opinion, and Disclaimer Opinion.

2.4. Public Welfare

The public welfare is one of the keys to the creation of good governance. Civil society reflects a prosperous and just society. Some parameters of civil society are the level of human development, the level of society knowledge, the level of life expectancy, and the level of welfare. Civil society has a better perspective in seeing its government so that it can create a better oversight and governance function, both in terms of economic, social and cultural aspects (Despotis, 2005).

One of the parameters in measuring the welfare of the societies that is manifested in civil society is the Human Development Index (HDI) of each region (Istiyanto, 2016). One of the parameters in the level of welfare is human development index in which there is a public health index measured from the average life expectancy of the public. The higher life expectancy, the better the level of public health (Adzani & Martani, 2014). The level of welfare could also be measured by public’s per capita income. Giroux and McLelland (2003) prove that the level of public income influences audit quality and the level of local government financial management.

2.5. Audit Findings

Findings are the material issues found by an auditor during the audit of the financial reports and these issues should be presented and communicated with the audited entity to make improvements and performance improvements in maintaining the quality of the financial reports. Audit findings in this study incorporate the weaknesses of internal control system findings and non-compliance with legislations requirements findings.

Weaknesses of internal control system are grouped into three, namely, (1) weaknesses in accounting and reporting control systems, (2) weaknesses of control system of budgeting of revenue and expenditure, and (3) weaknesses of internal control structure. Meanwhile, the findings of non-compliance with legislations consists of (1) compliance, (2) regional losses, (3) potential loss of the area, (4) lack of acceptance, and (5) administrative deviations.

2.6. Follow-Up Audit Recommendations

Article 20 of the Act Number 15 of 2004 stated that officials should follow up the recommendations through an examination. The Audit Board monitors the follow-up of the results, and informs the House of Representative. Article 6 Section 1 Government Regulation Number 2 year 2017 concerning Monitoring Implementation of Follow-up Recommendations on the Examination of Audit Board mentions that the Audit Board reviews the answers or explanations received from the officials to determine whether the follow-up has been conducted in accordance with the recommendations of Audit Board. In paragraph 4, the review results are classified in accordance with recommendations: follow-up is not in accordance with recommendations, recommendations have not been followed up, and recommendations cannot be followed up. Financial reports are the form of accountability for state or local financial management during a period. Based on Article 5 of Governments Regulation Number 8 of 2006 on Financial Reporting and Performances of Government Agencies, the State Financial Institution/Institution/Regional Work Unit consists of (1) Budget Realization Report; (2) Financial Position; (3) Cash Flow Statements; and (4) Notes to Financial reports (CaLK).

In 2015, the accrual-based accounting was adopted by the local government to complete the implementation of Government Regulation Number 71 of 2010, which explains the Government Accounting Standards (GAS). Based on these regulations, LGFR contributed three reports, namely, Changes in the Remaining Budget (LPSAL), Reports Operational (LO), and Equity Change Report (LPE) (IHPS I, 2016).

3. Hypothesis Development

H1a: Audit findings have direct negative effect on audit opinion.

H1b: Audit findings have indirect negative effect on audit opinion through disclosure level of provincial government financial reports.

H2a: Follow-up audit recommendations have direct positive effect on audit opinion.

H2b: Follow-up audit recommendations have indirect positive effect on audit opinion through disclosure level of provincial government financial reports.

H3a: Human Development Index (HDI) has a direct positive effect on audit opinion.

H3b: Human Development Index (HDI) has an indirect positive effect on audit opinion through disclosure level of provincial government financial reports.

H4a: The level of public health has a direct positive effect on audit opinion.

H4b: The level of public health has an indirect positive effect on audit opinion through disclosure level of provincial government financial reports.

H5a: The level of public income has a direct positive effect on audit opinion.

H5b: The level of public income has an indirect positive effect on audit opinion through disclosure level of provincial government financial reports.

H6: Disclosure level of provincial government financial reports has a direct positive effect on audit opinion.

The research model of the study is as follows:

4. Research Methods

The data that used in this research is secondary data. The data were taken from the websites of Badan Pemeriksa Keuangan Republik Indonesia (https://www.bpk.go.id/) and Badan Pusat Statistik (https://www.bps.go.id/). The object of this research were Provincial Government Financial Reports in Indonesia examined by the Audit Board for fiscal years 2016 to 2018, which amounted to 102 financial reports. Based on the purposive sampling criteria, 84 financial reports fulfilled the criteria and 18 financial reports did not fulfil the criteria. Procedure of sample selection is presented in Table 1.

Table 1: Procedure of Sample Selection

In this study, the hypothesis testing tool is path analysis using the Statistical Package for Social Sciences (SPSS) version 15. Each line between variables has a path coefficient to measure the influence of each independent variables on the dependent variable. The path coefficient value is calculated by using multiple regression analysis. To answer the questions and simultaneously test the hypotheses, the empirical model in this research is formulated as follow:

\(Model 1 \\ \begin{aligned} \mathrm{OPINION}=& \rho_{1} \mathrm{FIND}+\rho_{2} \mathrm{FAR}+\rho_{3} \mathrm{HDI}+\rho_{4} \mathrm{HEALTH} \\ &+\rho_{5} \mathrm{INCOME}+\rho_{6} \mathrm{DISC}+e \\ &(\mathrm{H} 1 \mathrm{a}, \mathrm{H} 2 \mathrm{a}, \mathrm{H} 3 \mathrm{a}, \mathrm{H} 4 \mathrm{a}, \mathrm{H} 5 \mathrm{a}, \mathrm{H} 6) \end{aligned} \)

Figure 1: Research Model

\(Model 2 \\ \begin{aligned} \mathrm{DISC}=& \rho_{7} \mathrm{FIND}+\rho_{8} \mathrm{FAR}+\rho_{9} \mathrm{HDI}+\rho_{10} \mathrm{HEALTH} \\ &+\rho_{11} \mathrm{INCOME}+e \end{aligned}\)

Index:

FIND: Audit Findings

FAR: Follow-up of Audit Recommendations

HDI: Human Development Index (HDI)

HEALTH: Public Health

INCOME: Public Income

DISC: Disclosure Level of Financial Reports

OPINION: Audit Opinion

e: Error

Audit opinion is measured using an ordinal scale from the highest ranking to the lowest ranking of opinion because the highest ranking is considered the best of audit opinion. Audit opinion sorted according to rank consists of 4 for Unqualified Opinion including Modified Unqualified Opinion, 3 for Qualified Opinion, 2 for Adverse Opinion, and 1 for Disclaimer of Opinion. Audit opinion measurement refers to the research by Sari et al. (2015). Audit findings in this study is measured based on the natural logarithm of total nominal value of weaknesses internal control system findings plus non-compliance with legislation findings based on the Audit Board Examination Report (LHP) of the Provincial Government Financial Reports. Audit findings measurement refers to the research by Sari et al. (2015).

Follow-up audit recommendations in this study are measured based on the total nominal of Audit Board’s recommendations that are followed up plus recommendations that cannot be followed up divided by the total nominal of recommendations. Follow-up audit recommendations measurement refers to the research by Sari et al. (2015). The data used is seen from the value of human development index in all provinces of Indonesia. HDI measurement refers to the research by Surya and Suparno (2019). The measurement of the level of public health in this study is based on the average life expectancy of the population. The level of public health measurement refers to the research by Adzani and Martani (2014). The level of public income measured by the natural logarithm of GRDP per capita of each province. The level of public income measurement refers to the research by Istiyanto (2016).

The intervening variable in this research is the disclosure level of provincial government financial reports by using a comparison between the disclosures that have been presented in the CaLK based on a checklist in accordance with Government Regulation Number 71 of 2010 concerning Government Accounting Standards (GAS) (Sari et al., 2015). Disclosure level of provincial government financial reports can use a formula:

\(\mathrm{DISC}=\frac{\text { disclosure in provincial government financial repots }}{\text { disclosure in GAS }}\)

5. Results and Discussion

5.1. Regression Results

The regression result for model 1 is shown in Table 2 and for model 2 is shown in Table 3.

Table 2: T-Test Model 1

a. Dependent Variable: OPINION.

Table 3: T-Test Model 2

a. Dependent Variable: DISC.

Based on Table 2, the first regression model is:

\(\begin{aligned} \mathrm{OPINION}=&(-0.212) \mathrm{FIND}+(-0.034) \mathrm{FAR} \\ &+(-0.288) \mathrm{HDI}+0.122 \mathrm{HEALTH} \\ &+0.252 \mathrm{INCOME}+0.133 \mathrm{DISC}+e \end{aligned}\)

The result of hypotheses test in this research is, as follow:

5.1.1. Audit Findings on Audit Opinion

Table 2 shows that audit findings variable have a negative regression coefficient (Beta) value of –0.212 and the Sig. value of 0.048 < α (0.05). It means that audit findings negatively and significant affect audit opinion. Thus, the first hypothesis (H1a) that states that audit findings have a negative effect on audit opinion is accepted. This result is in accordance with the research conducted by Kusumawati and Ratmono (2017), which found a negative influence of audit findings on audit opinion.

The direction of negative correlation indicates that the higher the audit findings, the worse the opinion by the provincial government financial reports from the Audit Board. It means that the smaller the number of audit findings related to weaknesses of internal control systems findings and non-compliance with legislations findings, the better the opinion by the provincial government. Therefore, if the number of audit findings is higher, it can encourage the provincial government to decrease the findings in the next period and the financial reports will obtain an unqualified opinion.

5.1.2. Follow-up of Audit Recommendations on Audit Opinion

Table 2 shows that follow-up audit recommendations variable have a negative regression coefficient (Beta) value of –0.034 and the Sig. value of 0, 771 > α (0, 05). It means that follow-up audit recommendations do not significantly affect audit opinion. Thus, the second hypothesis (H2a) that states that follow-up audit recommendations have a positive effect on audit opinion is rejected.

This result is not in accordance with the research conducted by Din et al. (2017), which found that follow-up financial investigation on audit opinion has a significant positive effect. If the recommendation of the audit result is completed by the higher audit, the findings related to the previous period’s loss will be expected not to be diminishing and the opportunity of local government to get an unqualified opinion is also higher (Setyaningrum et al., 2013).

This research does not support the signaling theory, which stated that the provincial government would maintain good audit opinion from the previous period by increasing the percentage of follow-up of audit recommendations. Data shows that the number of audit recommendations of several provinces is not followed up optimally by the government. Therefore, the number of audit recommendations does not encourage the provincial government to improve the follow-up audit recommendations from the Audit Board and improve the quality of audit opinion.

5.1.3. Human Development Index on Audit Opinion

Table 2 shows that human development index variable has a negative regression coefficient (Beta) value of –0.288 and the Sig. value of 0.124 > α (0.05). It means that human development index do not significantly affect audit opinion. Thus, the third hypothesis (H₃a) that states that human development index has a positive effect on audit opinion is rejected.

The result research by Istiyanto (2016) found that human development index has no significant effect on audit opinion of LGFRs. This is due to the fact that opinions from the Audit Board are given to provincial government without the need to consider whether the public in the area have a high or low level of human development index. This is in line with the Standar Pemeriksaan Keuangan Negara (SPKN/State Financial Audit Standards), which states that the examination of local or provincial government financial reports is a type of financial audit by the Audit Board with the aim of giving opinion statements about the level of fairness of the information presented in local or provincial government financial reports.

Therefore, the higher or lower level of human development index in the province will not affect the audit opinion of provincial government financial reports. This is due to the aim of the Audit Board giving opinion statements about the level of fairness of the information presented in provincial government financial reports without considering whether the public in a province has a high or low level of human development index.

5.1.4. Public Health on Audit Opinion

Table 2 shows that public health variable has a positive regression coefficient (Beta) value of 0.122 and the Sig. value of 0.552 > α (0.05). It means that public health does not significantly affect audit opinion. Thus, the fourth hypothesis (H4a) that states public health has an effect and positive on audit opinion is rejected. This result is in line with the research by Adzani and Martani (2014), which found that the level of public health did not significantly influence local government audit opinion.

One of the reasons public health did not significantly influence provincial government audit opinion is that there is an anomaly when seeing trends in Indonesia. The anomaly occurred in Bengkulu, DKI Jakarta, Kepulauan Bangka Belitung, North Maluku and Maluku where the provincial government financial reports have a bad audit opinion, but the life expectancy of the most society is actually relatively higher compared with West Nusa Tenggara, West Papua and West Sulawesi where the society’s life expectancy is relatively lower, but the provincial government financial reports have a good audit opinion.

Therefore, the higher or lower level of life expectancy in the province will not affect the audit opinion of provincial government financial reports. This is due to the aim of the Audit Board giving opinion statements about the level of fairness of the information presented in provincial government financial reports without considering whether the public in a province has a high or low level of life expectancy.

5.1.5. Public Income on Audit Opinion

Table 2 shows that public income variable has positive regression coefficient (Beta) value of 0.252 and the Sig. value of 0.048 < α (0.05). It means that public income has positively significant effect on audit opinion. Thus, the fifth hypothesis (H5a) that states that public income has a positive effect on audit opinion is accepted. This result is in line with the research by Adzani and Martani (2014) and Surya and Suparno (2019), which found that the income per capita has a positive influence on the audit opinion of local government financial reports.

This shows that the more prosperous the society is, the easier the society strives to fulfill its primary and secondary needs so that the society is more independent, active and critical in overseeing government financial managament (Adzani & Martani, 2014). This is consistent with the research by Giroux and McLelland (2003), which proves that the level of public income has a positive correlation toward local government governance.

Therefore, this indicates that when people’s income increases, it no longer strives to fulfill its primary needs and more concerned with the government so that supervision from the society will increase also on the government financial management in the use of financial resources to be more in the interests of the society.

5.1.6. Disclosure Level of Financial Reports on Audit Opinion

Table 2 shows that disclosure level of financial reports variable has regression coefficient (Beta) value of 0.133 and the Sig. value of 0.026 < α (0.05). It means that disclosure level of financial reports has positively significant effect on audit opinion. Thus, the sixth hypothesis (H₆) that states that disclosure level of financial reports has a positive effect on audit opinion is accepted. This result is in line with the research by Priharjanto and Wardani (2016), which found that audit opinion positively and significantly affects the disclosure level of provincial government financial reports.

The direction of positive correlation indicates that the better opinion obtained, the higher the disclosure level of financial reports produced, indicating the quality of financial reports is good. This is in accordance with one of the criteria for granting an audit opinion, which is the adequacy of disclosure. In addition, management assertion is considered in assessing the reasonableness of financial reports, which is the assertion of completeness and presentation as well as disclosure.

Pursuant to the Act Number 15 of 2004 concerning the Audit of State Financial Management and Responsibility Opinion, it is based on criteria, namely, conformity to Government Accounting Standards (GAS), adequacy of disclosure, compliance with laws and regulations, and effectiveness of internal control system.

Moreover, the research results of Naopal et al. (2017) also found that audit opinion positively and significantly affects the disclosure level of local government financial reports. It means that increasing the disclosure level of information will increase the quality of audit opinion. It proves that the better the audit opinion, the higher the disclosure level of financial reports.

According to agency theory, the government as an agent has direct access to information compared to the society, which leads to information asymmetry. To reduce this conflict, society as principal needs to monitor government performance. Therefore, the government is obligated to issue financial reports as a form of responsibility and increased accountability and transparency in an attempt to reduce information asymmetry.

5.2. Path Analysis and Discussion

5.2.1. The Influence of Audit Findings on Audit Opinion through Disclosure Level of Financial Reports

The significant value of the direct effect of audit findings on audit opinion is 0.048 < 0.05 and disclosure level of financial reports on audit opinion is 0.026 < 0.05 as well as audit findings on disclosure level of financial reports is 0.040 < 0.05. It can be concluded that the disclosure level of financial reports did not mediate the audit findings on audit opinion. Thus, the first hypothesis (H1b) was rejected.

The result of this research is in line with the research by Hilmi and Martani (2012), which found that audit findings do not significantly influence the disclosure level of provincial government financial reports. The number of audit findings on audit provincial government financial reports does not significantly influence the disclosure level of financial reports. The audit findings from the Audit Board do not encourage the provincial government to provide the level of disclosures becomes increasingly higher.

Hendriyani and Tahar (2015) also stated that a large number of audit findings did not affect the disclosure because of the Audit Board would provide recommendations to decrease the findings, with an improvement the financial reports will obtain an unqualified opinion, thus that the number of audit findings did not affect the disclosure of financial reports.

Therefore, if many infringements were found in the provincial government financial reports (audit findings), it will not affect the disclosure of financial reports. This is because the Audit Board will provide recommendations to the government in order to fix the audit findings and subsequently, with the improvement carried out, the provincial government financial reports will obtain an unqualified opinion. This suggests that there needs to be a strong commitment and efforts from provincial government to implement the recommendations.

5.2.2. The Influence of Follow-up of Audit Recommendations on Audit Opinion through Disclosure Level of Financial Reports

The significant value of the direct effect of follow-up audit recommendations on audit opinion is 0.771 > 0.05 and disclosure level of financial reports on audit opinion is 0.026 < 0.05 as well as follow-up audit recommendations on disclosure level of financial reports is 0.997 > 0.05. It can be concluded that the disclosure level of financial reports mediates the follow-up audit recommendations on audit opinion. Thus, the second hypothesis (H2b) was accepted. The research conducted by Sari et al. (2015) found the follow-up audit recommendations on audit opinion through disclosure level of local government financial reports had a positive effect.

The direction of the positive correlation indicates that the more recommendations from the Audit Board that are followed up according to the recommendations will improve the quality of the provincial government financial reports represented by a high level of disclosure. The higher the disclosure level of financial reports, the better the probability of opinion obtained by the provincial government.

In view of signaling theory, governments will give signals to the public by providing quality of financial reports, improving internal control systems, more complete disclosures and more detailed explanations in disclosures. The government can also package more information on achievement and financial performance in order to demonstrate that the government has carried out the mandate given by the public (Hendriyani & Tahar, 2015).

5.2.3. The Influence of Human Development Index on Audit Opinion through Disclosure Level of Financial Reports

The significant value of the direct effect of human development index on audit opinion is 0.124 > 0.05 and disclosure level of financial reports on audit opinion is 0.026 < 0.05 as well as human development index on disclosure level of financial reports is 0.031 < 0.05. It can be concluded that the disclosure level of financial reports did not mediate human development index on audit opinion. Thus, the third hypothesis (H3b) was rejected. The research conducted by Najah and Purwati (2019) found that human development index did not significantly influence the disclosure level of local government financial reports.

Data shows that the higher level of human development index in an area is not followed by the disclosure level of financial reports made by regional governments. Therefore, this shows that society as principal has a less active role in encouraging regional governments as agents to provide more detailed disclosures and causing regional government to not be motivated to increase disclosures in their financial reports. In addition, Istiyanto (2016) revealed the fact that opinions from the Audit Board are given to provincial government without the need to consider whether the public in an area has a high or low level of human development index.

5.2.4. The Influence of Public Health on Audit Opinion through Disclosure Level of Financial Reports

The significant value of the direct effect of public health on audit opinion is 0.552 > 0.05 and disclosure level of financial reports on audit opinion is 0.026 < 0.05 as well as public health on disclosure level of financial reports is 0.231 > 0.05. It can be seen that the significant value of public health on audit opinion > 0.05 and disclosure level of financial reports on audit opinion < 0.05 as well as public health on disclosure level of financial reports > 0.05, leads to the conclusion that the disclosure level of financial reports mediate public health on audit opinion. Thus, the fourth hypothesis (H4b) was accepted. This result is in line with the research by Ramachandran (2002), which revealed that society participation in government supervision is based on the level of fulfillment of needs and level of welfare. One of parameters in the level of welfare is human development index in which there is a public health index.

One of the parameters in the level of welfare is human development index in which there is a public health index measured from the average life expectancy of the public. The higher life expectancy shows the better level of public health (Adzani & Martani, 2014). The higher life expectancy of the society in a province, the more diverse the society’s needs will be fulfilled, which leads to society demands on the government in order that the government provides more complete and detailed disclosure of the provincial government financial reports. Thereby, if the disclosure level of provincial government financial reports is high, the probability of provincial government financial reports getting an unqualified opinion will be even higher.

5.2.5. The Influence of Public Income on Audit Opinion through Disclosure Level of Financial Reports

The significant value of the direct effect of public income on audit opinion is 0.048 < 0.05 and disclosure level of financial reports on audit opinion is 0.026 < 0.05 as well as public income on disclosure level of financial reports is 0.045 < 0.05. It can be concluded that the disclosure level of financial reports did not mediate public income on audit opinion. Thus, the fifth hypothesis (H₅b) was rejected. This research is in line with the research by Istiyanto (2016), which proves that the level of public income does not affect the audit opinion of local government financial reports. This is due to the fact that the level of public income that is the scope of the audit examination is only limited to the income presented on the financial reports and does not consider whether the level of public income are high or low.

Nor et al. (2019) found that the level of public income did not significantly affect the disclosure level of local government financial reports websites or the higher level of the public income in an area does not encourage the government’s desire to disclose financial reports on local government websites. Therefore, the level of public income does not affect the audit opinion through disclosure level of provincial government financial reports. This is due to the fact that the level of public income per capita that is the scope of the audit examination is only limited to the income presented on the financial reports and does not consider whether the level of public income in an area is high or low. In addition, the higher level of public income in an area also does not encourage the local or provincial government’s desire to improve the disclosure.

6. Conclusion

This research concludes that the audit findings, public income, and disclosure level of financial reports significantly influence audit opinion. Follow-up audit recommendations and the level of public health significantly influence audit opinion through disclosure level of provincial government financial reports in Indonesia.

The implications of this research are that the Provincial Government in Indonesia should be motivated to decrease the audit findings, increase the disclosures, and improve the provincial government efforts to implement the recommendations given by the Audit Board. The higher income per capita and life expectancy of society will result in higher monitoring by society in overseeing government financial management such as accountability of financial reports so that the pressure on provincial government to provide more information to the society is increasingly greater.

Future studies should combine secondary data and primary data through interviews to strengthen the research analysis on the factors that encourage provincial government to improve the financial reports quality, increase follow- up audit recommendations, and provincial government efforts to reduce audit findings. In addition, future studies could consider another variable such as the quality of internal auditors by an index based on the length of the assignment, training followed and educational background (Setyaningrum, 2015).