Internal audit function, political characteristics, and compliance with government accrual-based accounting standards in Indonesian local government

We investigate the effect of local government internal audit function and political characteristics on accounting compliance based on government accrual-based accounting standards in Indonesia while controlling for financial and local government characteristics. Our analysis is conducted for 2015-2018 with panel data regression, equipped with a robustness test using logistic regression. Our results show that internal audit function capability, maturity, expertise


Introduction
Indonesia has been working on improving government financial accountability for the past two decades.
Since the implementation of regional autonomy in 2002, Indonesian local governments have managed their financial resources. Consequently, they are required to prepare a Local Government Financial Statement (LKPD) to report their financial management (Halim & Abdullah, 2006). The preparation of LKPD, in the process, will be highly affected by the specified accounting basis and standards (Muda et al., 2018). In this case, the implementation of accrual accounting standards has been introduced to public sector institutions in many countries for the last two decades to realize New Public Management (NPM) (Baker & Rennie, 2006;Christensen & Parker, 2010).
It is generally accepted that accrual-based accounting can present more reliable financial information and improve financial statement quality and government accountability accordingly (Mir & Sutiyono, 2013). Therefore, Indonesia has also started to practice accrual-based accounting for local 27 governments regulated in the PP/71/2010 1 and Permendagri/64/2013 2 . The accrual-based accounting and the accrual-based Government Accounting Standards (SAP) started in 2015, replacing the implementation of cash basis accounting that they previously used. However, recent data from the Indonesia Supreme Audit Board (BPK) indicates that financial reporting in the Indonesian public sector still experiences several issues subject to accrual-based accounting implementation.
The audit opinion on the financial statement can be one of the indicators for successful accounting practice. For the past five years, several local governments could still not obtain unqualified audit opinions for their financial statements. As the external auditor of reports, the Indonesia Supreme Audit Board (BPK) reports 211 cases in 2015, 158 cases in 2016, 127 cases in 2017, and 97 cases in 2018 of local government inability to prepare a financial statement that obtained unqualified audit opinion. In addition to facing various financial management issues, these local governments cannot perform accounting processes for certain accounts such as assets, revenues, expenditures, and other accounts properly as regulated in the accrual-based SAP. Therefore, the implementation of local government accrual-based SAP in Indonesia deserves a deeper investigation, primarily due to the lack of empirical studies on this issue in Indonesian literature.
The level of accrual-based accounting implementation in the Indonesian local government, according to Wahyudi (2019), is varied in terms of quality. Several local governments have successful implementation while several others also struggle to follow. To realize effective implementation, every country needs to institutionalize the practice of accrual-based accounting (Harun, Van Peursem, & Eggleton, 2012). In the context of Indonesian government institutions, the internal auditor can be considered essential to realize successful accrual-based accounting implementation thanks to its assistance and monitoring function as mandated in PP/60/2008 3 . For Indonesian local governments, the internal audit function development adopts the concept from COSO (Winarna, Muhtar, Sutaryo & Amidjaya, 2021) in which the duty is performed by the local government inspectorate. The local government inspectorate carries out several important tasks such as ensuring financial processes efficiency, asset safety, compliance regulations, and financial statement reliability. Concerning financial reporting, the local government inspectorate plays an essential role in supervising and reviewing every financial reporting process in all reporting entities within a local government.
The external supervision of local government financial management also needs significant attention, especially from political factors (Bourdeaux & Chikoto, 2008). As a democratic country, Indonesia significantly relies on legislative supervision so that the executive can run local government properly. Legislative supervision characteristics have been evidenced to hold an essential part in local government financial accountability. Cohen & Leventis (2013) reveal that the political composition of the legislative does matter in realizing proper supervision on local government financial reporting. Araujo & Tejedo-Romero (2016b) highlight the influence of legislative gender factor in which the presence of women councillors may enhance the supervision process into more detailed and accurate. The size of the legislative board is also an essential factor for local government budget implementation in delivering public service (Lewis, 2019). Das (2012) emphasizes the crucial role of internal audit function in financial reporting for the government. In terms of empirical evidence, Chalu (2019) and Teruki, Nyamori, & Ahmed (2019) find a positive association between internal audit function and financial report quality. Further investigation on internal audit function characteristics such as personal traits (e.g., independence, skills, and capability) (Abbott, Daugherty, Parker, & Peters, 2016;Kusnadi, Leong, Suwardy, & Wang, 2016), experience, and specialization (Nogueira & Jorge, 2017), effectiveness (Ahmed Haji & Anifowose, 2016), as well as conflict management, and ethics (Thomas & Purcell, 2019) also demonstrate a positive effect. In Indonesia, the role of internal audit function in local government financial reporting is not widely explored yet, especially in realizing financial report compliance with government accrual-accounting standards applied from 2015. Thus, an investigation on local government internal audit function factors becomes very important. To fit the Indonesian setting, we draw several factors that consist of capability, maturity, expertise, and size as these characteristics have become the basis for local government internal audit function development.
In addition to internal audit function and political supervision on local government financial report preparation, it is also essential to consider the nature of the local government itself. In this study, we take local government financial factors as well as administrative factors into account. The financial resources managed by local governments also matter. Thus, we pick major financial characteristics such as total assets, revenue, and expenditure for our investigation as these are very decisive in local government financial management. As for administrative factors, we consider the significance of local government complexity, council type, and geographic location as these factors may share the gap among local governments.
This research provides several significances to fill the gap in the literature. We examine all local governments in Indonesia in order to answer the issue of limited local government observations in previous studies (Gamayuni, 2018;Miraj & Wang, 2019;Muda et al., 2018;Pamungkas, 2018;Suparman, Siti-Nabiha, & Phua, 2015). High coverage will provide a higher generalization ability of the research. In order to answer subjectivity issue due to the use of a survey in prior studies, we make use of official data obtained from the authorized institution, namely: (a) Indonesian local government list from the Ministry 29 of Internal Affairs; (b) local government accrual-based accounting compliance, financial characteristics, and profile from the Supreme Audit Board (BPK); and (c) local government internal audit function assessment result from the Financial and Development Supervisory Agency (BPKP). With these significances, this research emerges as one of the earliest pieces of literature to provide a comprehensive analysis in the Indonesian context. As for the practical contribution, by revealing the role of internal audit function in accrual accounting compliance, this study is expected to provide recommendations for Indonesian accrual-based accounting practices, especially for the authorized institutions in performing their functions.

Agency theory
Introduced by Jensen & Meckling (1976), scholars have prevalent agency theory for its principal-agent relationship and agency conflict. The theory is very applicable in any context, including local government (Lane, 2005). In general, the local government acts as the public agent to deliver public services and social welfare in which the public delegates the authority to the local government to manage public economic resources. Halim & Abdullah (2006) mention three relationships in the Indonesian context in local government financial management: the relationship between public and legislative, legislative and executive, executive (regional head), and working unit heads. As represented by the legislative as the principal, the public demands transparency and accountability for financial management. Thus, the local government needs to prepare it in a financial statement.
Further, the financial statement will be audited by the Supreme Audit Board (BPK) as the external auditor to be given an audit opinion. Jensen & Meckling (1976) mentions that the agent needs to be monitored to ensure best practice. Internal audit function plays a crucial role in supervising financial report preparation to meet the accounting standards. Local governments are required to prepare a financial statement as mandated in UU/17/2003 4 . As for the standards, local government financial statements must comply with accrual-based accounting standards (SAP).

Accrual-based accounting standards implementation in Indonesian government
To achieve best practice in state financial management, Indonesia set to practice accrual-based accounting for government institutions. Accrual-based accounting can present more reliable financial information and improve financial statement quality and government accountability accordingly (Mir & Sutiyono, 2013).
The commitment to implementing full accrual basis in local government accounting practice is manifested in PP/71/2010 5 and Permendagri 64/2013 6 . In the process, accrual-based accounting practice in the Indonesian government is implemented step by step. The implementation began with cash toward accrual accounting implementation until the 2014 fiscal year. Then, from 2015 and on, full accrual-based accounting practice becomes mandatory for all government institutions under the Accrual-Based Government Accounting Standard (SAP). With full accrual-based accounting standards, it is expected to deliver more benefit for the user of financial information as well as various stakeholders. This practice is also a form of cost-benefit maximization from reporting cost.

Internal audit function characteristics on accrual-based accounting standards compliance
Indonesia pays great attention to government financial management supervision to ensure that public funds are well-managed for central and regional levels regulated in the PP/60/2008 7 concerning the Government Internal Control System that is developed based on COSO concept. This becomes one of the most important part and should be one of the priorities in in public governance reform (Urbanik, 2016). It is expected that government institution will be able to realize good operational efficiency, assets safety, financial report reliability, and compliance with the regulation (Winarna, et al., 2021). The element of government internal control system consists of control environment, risk assessment, control activities, information and communication, and monitoring (Benedek, Szenténé & Beres, 2014). Achieving good operational efficiency, assets safety, and compliance to regulations can be supported by having adequate control environment and activities equipped with proper risk assessment. Further, the government can also realize financial report reliability as the output of financial management. All of these goals realization will also be possible with proper information and communication as well as continuous monitoring.  Therefore, with government internal control system implementation, the supervision on government institution can be optimized to realize the desired organizational goals.
In the context of local government, the duty of internal supervision is performed by the local government inspectorate with its internal auditor apparatus formally called the Government Internal Oversight Apparatus (APIP). The development of internal auditors in the Indonesian local government also becomes a significant concern. In this regard, the Financial and Development Supervisory Agency (BPKP) is responsible for internal auditor development all over Indonesia. BPKP has issued several policies and regulations in order to organize and develop the function of an internal auditor. Several critical aspects of internal audit function are highly regulated, such as capability and maturity. Internal auditor expertise and human resource availability are also essential to realize proper financial management supervision in local government. Following the context of Indonesia, we draw four variables, namely: capability (Alzeban & Gwilliam, 2014;Institute of Internal Auditors, 2009;van Rensburg & Coetzee, 2016;Sari, Hastuti, & Tannar, 2019;Yusof, Haron, & Ismail, 2018), maturity (Chen, Chung, Peters, & Wynn, 2017;Mansour, 2010;Sarens, Allegrini, D'Onza, & Melville, 2011), expertise (Abbott et al., 2016;Alzeban & Gwilliam, 2014;Gras-Gil et al., 2012), and size (Abbott et al., 2016;Alzeban & Gwilliam, 2014;Oussii & Taktak, 2018;Sarens, et al., 2011). Our hypotheses are formulated as follows.

Internal audit function capability
Indonesia considers internal auditor capability as a crucial aspect as regulated in the Perka BPKP No.
PER-1633/JF/2011 8 . Indonesia aims to continuously improve and monitor APIP capability by adopting the Internal Audit Capability Model (IACM) developed by the Institute of Internal Auditors (IIA). Here, IACM provides a structured framework for APIP development. Based on the assessment score, there are five levels in IACM to illustrate the capability of the internal audit function in performing their role that ranges from level 1 to level 5 (Initial, Infrastructure, Integrated, Managed, and Optimizing). Indonesia adopts this framework and conducts an annual assessment on APIP capability. With this assessment, BPKP can determine a set of recommendations to the local government inspectorate for continuous improvement. Several important elements can be improved: role and service, human resource management, professional practice, performance accountability and management, organizational culture and relationship, and governance structure. Therefore, the role of internal audit function will be maximized to improve the whole processes in local government finally.

32
In the internal audit literature, internal audit capability illustrates the ability of the internal audit unit to perform supervision on financial management (Alzeban & Gwilliam, 2014). The higher the capability, the higher the capability of the internal audit function in delivering its functions (Gras-Gil, Marin-Hernandez, & de Lema, 2012). The internal audit function will be more competent in providing assurance on governance, risk management, and internal control (Asare, 2009). Institute of Internal Auditors (2009) points that internal audit function capability will improve financial reporting and financial statement quality following specified accounting standards. Sawyer (2003) adds that internal audit function assessment will help enhance the accuracy and reliability of financial information. It is also revealed that a well-performing internal audit function will support the financial reporting process and its output quality (Alzeban, 2019). Miraj & Wang (2019) find that the difficulty of accrual-based accounting implementation is related to the lack of monitoring from internal audit function. Higher capability will help overcome this issue. Das (2012) supports the finding by revealing that the internal audit function plays a strategic role in accrual-based accounting standard implementation in a government institution.
Higher internal audit function capability is expected to support local government in preparing a financial statement following accrual-based accounting standards. Thus, we hypothesize that: H1: Internal audit function capability has a significant positive effect on government accrualbased accounting standards compliance

Internal audit function maturity
In addition to capability, the maturity of the government internal control system (SPIP) also becomes a significant concern for government financial management supervision in Indonesia 9 . Cooke-Davies (2004) states that the term maturity means optimally developed. The maturity concept plays an essential role in SPIP implementation by providing a structured and continuous development. Based on the assessment score, there are six local government SPIP maturity levels in Indonesia, ranging from level 0 to level 5 (Absent, Initiated, Developing, Defined, Managed and Measured, and Optimum). SPIP maturity level is very useful to evaluate and guide the implementation of SPIP. Abdolmohammadi, D'Onza, & Sarens (2016) posit that the maturity model helps internal audit function in improving its role in supervising financial processes for a better outcome in which higher maturity level can increase success line with the concept from the Institute of Internal Auditors (2014). Accordingly, it may help local governments to manage funds and finally deliver better public services properly.

33
Prior studies demonstrate that internal audit function maturity is necessary for assisting financial reporting activities (Chen, et al., 2017). Higher maturity allows internal audit function to engage in more comprehensive monitoring, and thus the accounting practice and financial report preparation will be able to meet the existing standards. Hepworth (2003) highlights the importance of internal audit function for local government accrual-based accounting practice, especially in its early stage due to the significant changes. Mansour (2010) further underlines the importance of internal audit function maturity in delivering support to the organization so that organization is able to adapt to regulation changes.
Accordingly, higher SPIP maturity for local government is expected to support the financial reporting processes to meet the requirements of government accrual-based accounting standards (SAP). Therefore, we formulated the following hypothesis: H2: Internal audit function maturity has a significant positive effect on government accrualbased accounting standards compliance

Internal audit function expertise
Expertise is a crucial characteristic for internal auditors to perform their duties. However, in Indonesia, it has been evidenced that government institutions lack accounting staff with the proper expertise, so accrual-based accounting practices in the public sector become ineffective (McLeod & Harun, 2014).
This circumstance results in lower accrual-based accounting implementation success both for central and local government levels. In the Indonesian setting, the internal auditor's functional position is classified into "skilled" and "expert" auditor. An inspectorate with more expert auditors indicates higher expertise. Abbott et al. (2016) mention the necessity of auditor expertise as a crucial attribute in realizing internal audit work quality. The more local government inspectorate staff with functional auditor position, the higher the possibility of better supervision function, including the financial reporting aspect. As a result, financial statement quality will be higher (Gras-Gil, et al., 2012). Qualified human resources in the internal audit function will have an implication on the information quality of financial statements (Lismawati, 2012;Toding & Wirakusuma, 2013). Therefore, it will result in better financial statement compliance on accrual-based accounting standards. With the support of the explanation above, we hypothesize that: H3: Internal audit function expertise has a significant positive effect on government accrualbased accounting standards compliance

Internal audit function size
The implementation of accrual-based accounting practice in Indonesian government institutions lacks functional staff that can perform proper accounting functions (McLeod & Harun, 2014). They argue that this circumstance underlies the low level of implementation success. This applies in every government from the top level down to the local government level. For the sake of a good financial reporting process, the local government should have an adequate number of internal to perform proper accounting practice, including those in inspectorate to perform the monitoring function. A local government with higher internal auditor availability is a step toward realizing internal audit function work quality (Abbott et al., 2016). A larger internal audit function size is related to its practice (Sarens, et al., 2011). The more local government inspectorate staff with functional auditor position, the higher the possibility of better supervision function, especially financial reporting. A higher number of qualified internal auditors is expected to realize adequate supervision so that financial statement quality can also be expected to be higher (Lismawati, 2012). Finally, the local government can produce a financial statement that complies with government accrual-based accounting standards. Therefore, we hypothesize that: H4: Internal audit function size has a significant positive effect on government accrual-based accounting standards compliance

Local government political factors on accrual-based accounting standards compliance
Political factors have also become very interesting in investigating local government accountability in Indonesia. Indonesia implements direct election for local government, both for legislative and executive election, since the reformation in the late 1990s. Legislative acts as people's representative to bring people's interest to the executive. Legislative also further perform supervision on executive actions in running the government. Therefore, the role of the legislative in supervising executive activities will be very crucial, considering several factors. Our study examines the role of legislative, political coalition, gender, and size.

Legislative political coalition
Political coalition becomes a big deal related to the supervision of inter-party activities and pressure to be more transparent (Araujo & Tejedo-Romero, 2016a;Cohen & Leventis, 2013). When political competition is low in which there are dominant parties, the legislative tends to be stronger so that the 35 supervision on local government financial management becomes lower. An executive with larger majority coalition support tends to shape higher opportunistic behaviour (Barnes & Jang, 2016). In contrast, the tighter political competition in which opposition is also strong, it can be expected to have a strict supervision on local government financial management (Araujo & Tejedo-Romero, 2016a;Caba Pérez, Rodríguez Bolívar, & López Hernández, 2014;Cohen & Leventis, 2013;García-Sánchez, Frías-Aceituno, & Rodríguez-Domínguez, 2013;García & García-García, 2010). With higher legislative supervision, the preparation of the local government financial statement will be more accurate and will be more likely to comply with government accrual-based accounting standards. Furthermore, Alcaide Muñoz, Rodríguez Bolívar, & López Hernández (2017) find that intense political competition is the primary driver of the level of transparency of government information. In addition, Berliner & Erlich (2015) also highlight the role of political supervision on local government institutional reform. Based on the findings above, we formulate the following hypothesis: H5: The legislative political coalition has a significant negative effect on government accrualbased accounting standards compliance

Legislative gender
The representation of women in politics also holds considerable contribution, according to previous studies. Araujo & Tejedo-Romero (2016b) reveal that women's political representation in municipalities positively increases information transparency and reduces information asymmetry. Women tend to have several characteristics that support better supervision of local government financial processes. Women are more likely to be more ethically minded than men (Fine, 2009). Women's stereotype also seems to be more favourable to increase communication and attempt to influence others and at the same time tends to be more collaborative and democratic (Eagly & Johannesen-Schmidt, 2001). Rodríguez-Garcia (2015) find that women are more responsive than men toward demand in local governance, manifested in better local government information disclosure. Accordingly, the higher presence of women in a legislative is expected to realize more effective supervision in local government financial management, especially in financial statement preparation to comply with government accrual-based accounting standards.
Therefore, we hypothesize that: H6: Legislative gender has a significant positive effect on government accrual-based accounting standards compliance

Legislative size
Another characteristic of legislation that becomes a significant concern in the literature is legislative board size (Lewis, 2019). The legislative members perform their function as people's representative and local government supervisor. Basically, a larger government requires more legislative members to perform its function (Britto & Fiorin, 2020). The larger number of legislative members may represent higher public interest as well as realize a more effective supervisory activity, and therefore, increases local government accountability. More legislators have a positive implication on local government financial management.
It can better control a budget maximizing bureaucracy (Pettersson-Lidbom, 2012) and increases government spending implementation (Egger & Koethenbuerger, 2010). With prior evidence aforementioned, it can be expected that larger legislative members will improve the supervision of accrual-based accounting standards in local government financial reporting. Therefore, the compliance will also be higher. Finally, we hypothesize that: H7: Legislative size has a significant positive effect on government accrual-based accounting standards compliance

Data collection and samples
This study aims to present a comprehensive examination covering all Indonesian local government at city/district level. Our analysis ranges from 2015 to 2018 fiscal year, in line with the implementation of accrual-based government accounting standards (SAP) that started in 2015 10 . The sample selection process is presented in Table 1.

37
Administratively, Indonesia has 514 district/city local governments in total. However, six (6) city local governments are not required to publish financial statements. These cities are those under Jakarta province in which their financial statements are consolidated to Jakarta province. Therefore, these cities are excluded from our observations. Finally, this research generates 2 032 observations, consisting of 508 district/city local governments for a four (4)

Variables and measurement
This study uses local government compliance with accrual-based SAP as the dependent variable. We investigate its determinants from three major aspects that later draw the independent variables. The first aspect comes from the local government internal audit function that consists of capability, maturity, expertise and size. The local government financial aspect is also important as it illustrates the object to be reported. We also control for financial and local government factors. Financial factors consist of total asset, revenue, and expenditure. We also pay attention to local government characteristics represented by the complexity, geography, and council type. The operationalization of research variables is presented in Table 2. LGTYPEit Dummy variable; 1 for city local governments, and 0 for district local governments Arifin, et al. (2015), Rakhman (2019) Source: Elaborated by the authors.
The measurement of local government accrual-based SAP compliance is referring to the BPK audit report. BPK audit report mentions specific accounts that do not comply with accrual standards, classified into ten (10) accounts: current assets, fixed assets, other fixed assets, long term investment, long term liability, revenue, operating expenditure, capital expenditure, operating expense, and other accounts.
As there are ten (10) accounts as indicators of non-compliance, each non-compliance account mentioned will reduce the compliance by 10%. Local governments that earn unqualified audit opinion will have a 100% accounting compliance level.
Local government internal audit function characteristics data are obtained from the BPKP assessment. The capability is measured with the IACM level, while the maturity is represented by the maturity score. Local government internal auditor expertise is measured by the proportion of expert auditor to the total auditor, whereas the size is represented by the total number of auditors in the local government internal audit function. As for political factor, legislative characteristics are employed to illustrate the political environment in a local government. The legislative coalition is measured by the proportion of legislative members support in the legislative board. Legislative gender also uses the proportion of women in the legislative board. Finally, the total number of legislative board members represents legislative size.
As for control variables, the financial factors that consist of local government total asset, revenue and expenditure are measured with its natural logarithm (Harumiati & Payamta, 2014;Sutaryo & Sinaga, 2018;Utama et al., 2019). As for local government characteristics, local government complexity is measured with a total number of working units (Adiputra, et al., 2018;Dalimunthe & Fadli, 2015;Maulana & Bestari, 2015). Local government geography is necessary as a control variable (Arifin, et al., 2015), measured with a dummy method; 1 for local governments located in Java and 0 for those located outside 40 Java following Rakhman (2019). Finally, local government type also uses dummy measurement; 1 for city local government and 0 for district local government (Arifin, et al., 2015;Rakhman, 2019).

Research model
Our study presents a thorough analysis that consists of descriptive statistics, correlation, panel data regression and robustness test. This research uses panel data regression analysis to test hypotheses. We choose the best estimation among common effect, fixed effect and random effect using the Chow test and Hausman test. The regression model is stated as follows: LGASSETit : Natural logarithm of local government total asset LGREVit : Natural logarithm of local government total revenue LGEXPit : Natural logarithm of local government total expenditure LGCOMit : Total number of local governments working unit LGGEOit : Local government geography LGTYPEit : Local government type it : Regression residual  Table 3 shows that the average local government compliance with accrual-based SAP for 2015-2018 has reached 91.9%. It can be said that the local governments have been able to practice accrualbased accounting pretty well the first four years of its implementation. The highest compliance and the mode are 100%, with a total of 1 439 observations. The lowest compliance level is 20%, with a 14.9% of standard deviation. There are 593 observations that have not fully complied with accrual-based SAP in which they still have several non-compliance accounts in their financial statement. Fortunately, the compliance trend keeps increasing, and non-compliance keeps decreasing throughout 2015-2018.

Descriptive statistics
However, the existence of 593 non-compliance cases may imply that there are still several difficulties faced by Indonesian local governments in implementing accrual-based accounting practice and SAP. The trend of local government accounting compliance with accrual-based SAP is presented in Figure 1. Further examination of 593 local governments that fail to reach full compliance reveals that most non-compliance cases are dominated by fixed asset account presentations with more than 80%. In addition to being a financial account with the highest amount in the financial statement, fixed assets account presentation requires a more complicated process and policy than other accounts such as beginning balance, depreciation, impairment, and revaluation. Local governments may be facing several difficulties in making policies related to fixed assets. Current assets account presentation emerges second, followed by capital expenditure, operating expenditure, operating expense, other fixed assets, long-term investment, other accounts, revenue, and lastly, long-term liability. Long-term liability account presentation has the least amount of non-compliance due to the several limitations in long-term financing policy for Indonesian local governments. Descriptive statistics results of internal audit function characteristics exhibit that the capability has an average value of 1.711, with the highest score of 3 and its lowest at 0. This implies that the average capability score of the Indonesian local government internal audit function based on IACM is in level 1 (initial) and is close to level 2 (infrastructure). As for the maturity, the average score for all local governments is 2.112 that is in level 2 (developing), with the highest maturity score at 3.

Correlations
We also test the correlation among accrual-based SAP compliance, internal audit function, financial and local government characteristics variables to examine the association between research variables. The result of our correlation test reveals that the variables from the internal audit function, political, financial characteristics exhibit positive correlation with the accrual-based SAP compliance. Meanwhile, local government complexity exhibits negative association with the accrual-based SAP compliance. The result of the correlation test is presented in Table 4 as follows. Notes: a : significant at 0.000 level; b : significant at 0.05 level; c : significant at 0.1 level Source: Secondary data, processed by the authors.

Panel data regression analysis
The primary analysis of our study aims to reveal the impact of local government internal audit function, political characteristics on accrual-based SAP controlled with financial and local government characteristics. We perform panel data regression along with the robustness test using STATA. To gain richer insights regarding the practice, we also divide our analysis based on the local government type, thus distinguishing the effect in the city and district local government. The results of Chow and Hausman tests demonstrate that the best regression estimation emerges with random effect testing. All of the regression scenario has good model fitness with positive value for the coefficient of determination and proper Chisquare statistics. The result is presented as follows: Overall, all internal audit function variables appear as significant predictors of accrual-based SAP compliance in our observation consists of the IAFCAPit, IAFMATit, IAFEXPit, IAFSIZEit. These results support the idea that the internal audit function will support the financial reporting process and its output quality (Alzeban, 2019). IAFCAPit demonstrates a positive effect. However, the effect is insignificant in city local government cases. Higher IAFCAPit will support the financial reporting process as internal auditors will be able to perform better supervision and control on local government, especially from budgeting process and budget implementation. With more comprehensive supervision, the accounting practice will be appropriately evaluated, resulting in financial statements that comply with accrual-based SAP. Gamayuni (2018) evidences a similar finding in which internal auditor capacity (a part of capability) positively affects financial statement quality in Indonesian local governments. As for the insignificant effect in city local government, we find that most city local governments have fully complied with accrual-based SAP (303 out of 372 observations). IAFCAPit in most cities is also above average at 1.869 and thus makes the effect insignificant.
IAFMATit demonstrates a strong positive effect in all testing scenario. With higher maturity, local government internal auditors will have structured processes and developed practices (DeSimone, D'Onza, & Sarens, 2019;Institute of Internal Auditors, 2014). This leads to consistent internal supervision and continuous evaluation to improve local government accounting practice (Chen et al., 2017). Therefore, the financial statement will be more likely to comply with accrual-based SAP. Our finding supports the idea from Mansour (2010), who mentions that higher organization maturity will be very helpful in facing new changes. It is also in line with the concept of organization maturity itself, that is, to help the organization achieve its goals more effectively (van Looy, de Backer, & Poels, 2011). IAFEXPit shows a significant positive effect except in the city local government examination panel. Following Abbott et al. (2016), more expert auditors in a local government, internal audit function are also needed for higher audit work quality. It is essential in conducting proper supervision and assistance on local government accounting practice. Therefore, local government will be more likely to have better information quality in the financial statement (Lismawati, 2012;Toding & Wirakusuma, 2013) and further realize good compliance on accrual-based SAP.
IAFSIZEit also exhibits a significant positive effect in all local government and city local government examination panel. We evidence that cities with higher accrual compliance tend to have a larger number of internal auditors. In practice, more staff is also essential for city local governments as they tend to have more working units compared to the district type. Thus, the internal audit function will require more individuals to conduct financial supervision and assistance. The low availability of competent human resources will also be a barrier for practicing accrual-based accounting in government institution (McLeod & Harun (2014).
Further, the availability of human resources in the internal audit function is necessary for city local government in regard to financial reporting, consistent with Mihret, James, & Mula (2010). More internal auditors may also have an implication on information quality (Lismawati, 2012;Toding & Wirakusuma, 2013). With all these settings, the local government can produce a financial statement that complies with accrual-based SAP.
All local government political factors investigated in this study also provide a significant positive effect. This finding is not in line with our hypothesis development based on prior studies in the literature. Instead of opening opportunistic behaviour for the reigning government, a higher political coalition may also lead to a better realization of local government activities as there is less intervention from the opposition. The circumstances possibly make local government programs implemented way easier and thus improving its financial accounting process. There is also a possibility that a higher political coalition supporting the incumbent regional head may support the reformation of local government state financial management. Therefore, a better accounting practice can be implemented and finally results in a good quality financial statement that complies with accrual-based SAP.
As for POLGENDit, our finding demonstrates a positive effect in line with Araujo & Tejedo-Romero (2016b). Women's representation in politics may reduce information asymmetry about local government activities. The natural characteristics of women for being more ethically minded than men (Fine, 2009), more collaborative and democratic (Eagly & Johannesen-Schmidt, 2001), as well as being more responsive (Rodríguez-Garcia, 2015) may help supervise local government financial processes, especially in financial statement preparation to comply with government accrual-based accounting standards. POLSIZEit also exhibits a positive contribution. The role of the legislative as people's representative is inevitable. A larger number of members may represent the higher public interest and realize a more effective supervisory activity in realizing local government accountability (Lewis, 2019).
This will improve the supervision of accrual-based accounting standard in local government financial reporting. Therefore, the compliance will also be higher.
As for financial characteristics, we find a significant effect in LGREVit with a positive direction.
Our result demonstrates that local governments with higher revenue are more likely to fully comply with 48 accrual-based SAP. Guillamón, Bastida, & Benito (2011) evidence that local governments with higher transfer and tax revenue generation provides better financial information and are more transparent. In this case, taxpayers demand high transparency in return for allowing governments to collect and manage large amounts of financial resources (Lassen, 2001). Moreover, local governments that can generate higher revenue are more developed, and thus they have better accrual-based accounting practice. Other financial factors that consist of LGASSETit and LGEXPit do not exhibit any significant effect.
Among the administrative characteristics, LGCOMit appears to be a significant predictor, with a negative effect on all observations and district government observations. The effect becomes insignificant in city local governments. This result implies that the complexity generally affects the financial reporting process, particularly in district type local governments. The higher the complexity, the less likely district local governments will achieve full compliance. Higher complexity, as represented by more working units, will result in more complex financial reporting. Every working unit as an accounting entity will be obliged to produce its financial statement, and all reports should be consolidated into one in the local government financial statement. In addition, the condition will require more financial supervision as well as consulting activities In which district local government may not be able to do. In city local governments, high complexity can be accommodated with larger internal audit staff so that financial reporting in all working units gets proper supervision and assistance. On the other hand, LGTYPEit and LGGEOit demonstrate no significant effect. We conduct a robustness test to check whether our research result is robust by performing logistic regression analysis. We give score 1 if the local government can fully comply with accrual-based SAP, while those who cannot fully comply with the standards will get 0 score. The result indicates that the effect of all variables remains consistent, followed by proper model fitness. We find little varied results in which the capability of an internal auditor and political coalition exhibit a positive effect in realizing full compliance of government accrual-based SAP. Other differences emerge from the local government revenue variable result that presents no significant impact in all examination scenarios. Local government assets also show a significant positive effect in the district local government result. Finally, we can state that our research model is robust and can deliver proper empirical evidence of accrual-based SAP compliance determinants from the internal audit function and financial and local government characteristics.

Conclusions
The implementation of accrual-based SAP in Indonesian local governments has not reached the maximum level. There is still an average of 29.1% of account presentation that do not comply with accrual-based SAP since the implementation started in 2015 up until 2018. Our study evidences a significant role of local government internal audit function in supporting accrual-based SAP. Internal audit function capability, maturity, expertise, and size positively affect the compliance level. With higher internal audit capability, maturity, and size, the supervision of local government accounting practices will be more 50 comprehensive and more likely to comply with accounting standards. We also reveal that political traits from the legislative board also matter in accrual-based SAP implementation thanks to the supervisory role. We find a positive effect of local government legislative political coalition, gender, and size, resulting in a higher compliance level.
Our findings provide several implications for Indonesian local governments and BPKP. For local governments, it is essential to improve the accounting practice by having proper accounting policies and competent human resources in accounting. For BPKP, as the authorized institution for local government internal auditor development, it is suggested that BPKP maximize their programs for internal auditor capability and maturity enhancement and expertise improvement. BPKP should also conduct a continuous evaluation to make the program implementation even better. With these achieved, the level of capability and maturity can be appropriately improved, and the need for internal audit human resource can be fulfilled. Furthermore, the legislative should also provide a proper political supervisory role as people's representative. Therefore, accrual-based SAP in Indonesian local governments can be appropriately implemented to accomplish greater public accountability. (Baltaci & Yilmaz, 2006).