Audit Committee Effectiveness on Earnings Management

This study aims to determine the effect of the number of audit committee meetings and the size of the audit committee on earnings management in manufacturing companies listed on the Indonesia Stock Exchange in 2016, 2017, 2018 and 2019. Earnings management is measured by Discretionary Accruals using the Modified Jones Model. The population in this study were 160 manufacturing companies listed on the Indonesia Stock Exchange for the period 2016-2019. The research data were obtained from the annual reports of manufacturing companies for the period 2016 to 2019. Based on the purposive sampling method, the sample obtained was 40 companies. The hypothesis in this study was tested using multiple linear regression analysis. The results of the test show that the number of meetings and size of the audit committee has a significant effect on earnings management .


Introduction
Financial reports are the final result of an accounting process and are a reflection of the condition of a company.Financial reports are also a medium for communicating financial information from management to parties outside management or companies who feel they have an interest in the company's financial information needs (Hariyanto & Suhardianto, 2018).Financial reports are expected to provide adequate information to stakeholders in taking a step or decision regarding the investment of funds they have.However, if there is a misstatement in the financial statements.Such information will be irrelevant for making a decision (Wahyuni & Budiwitjaksono, 2017).Information published by the company will be responded by the market.One of the information needed by investors is profit information generated in the company.Profit information can be used as one of the decisions for making a decision by investors.Profit information is often used as the target of engineering opportunistic actions by company management.In this case the company's management will act to benefit itself.So that it can often be detrimental to investors and creditors.The behavior of manipulating or managing company profits with specific goals in accordance with management's wishes is known as earnings management (Novi and Eli, 2018).Earnings management occurs because the management (agent) as the manager of the company knows more about the company's internal and prospects in the future compared to the company owner (principal).Knowledge of this information sometimes agents tend to convey information to owners that are not in accordance with the actual conditions of the company, so agents get opportunities to practice earnings management in order to maximize utility.(Scott, 2015:306).
(principal) hire other people (agent) to provide a service that delegates decision making to the agent.This agency theory assumes that each individual acts in their own interests (Lestari & Murtanto, 2017).Agency theory assumes that each party will prioritize their personal interests so that it will cause a conflict of interest between the agent and the principal.The agency relationship between the owner and the manager should produce a mutually beneficial relationship for all parties, but this will happen if each party carries out their rights and obligations responsibly.(Andanasari & Ayu 2017).

Earning Management
Earnings management is a factor that can reduce the reliability of financial reports (Wiryadi dan Nurzi, 2013).Healy and Wahlen (1999) in Iskak and Lulus (2016) states that earnings management is caused by management using judgment in compiling economic transactions so that the data in the financial statements changes.Management is able to predict future economic events with the policies they have made in the financial statements.This action will mislead stakeholders in making decisions.According to Wiryadi and Nurzi (2013) earnings management can be efficient (increase the profit information to be reported) and can also be opportunistic (management maximizes its personal interests).Earnings management is associated with the selection of accounting methods in achieving certain goals.
Audit committee meetings serve as a formal media for audit committee members in the context of oversight of the corporate governance process.Article FCGI (2002) mentions that the audit committee usually needs to hold meetings three to four times a year to carry out its obligations and responsibilities relating to the financial reporting system.Based on the Financial Services Authority Regulation Number 55/POJK.04/2015, the audit committee holds regular meetings at least 1 (one) times in 3 (three) month.In addition to conducting meetings with internal parties.The audit committee also holds executive meetings with parties outside the audit committee membership such as commissioners, senior management, head of internal auditors and head of external audit.
The audit committee is a committee formed by the company to carry out the supervisory process regarding the preparation of the company's financial statements, which aims to suppress an action that should not be carried out by company management.So that the company's financial statements presented can be trusted with the results of the information.The creation of an effective oversight function of the audit committee is related to the amount of resources that the committee has.The larger the size of the audit committee, the company will have sufficient resources to monitor the company (Widiastuty, 2016) The size of the audit committee is expected to improve the mechanism of checks and balances (supervision and alignment) within a company so as to minimize the practice of manipulating financial statements.
The frequency of regular meetings can reduce earnings management actions (Pamudji and Trihartati 2010).The audit committee is required to hold meetings 3 to 4 times a year to carry out obligations related to matters relating to the financial reporting process (Mintara 2008).The audit committee that meets less than the minimum number set, has the opportunity to re-occur earnings management (Abbott et al., 2005).

H1:
The number of audit committee meetings has an effect on earnings management.
An audit committee with a sufficient number of members will support the oversight function of the actions taken by agents.The audit committee consists of at least 3 members, both from independent commissioners and outsiders (Financial Services  The type of data used in this study is data derived from documents, namely the financial statements of sample companies.Data is obtained by documenting several items included in the financial statements.The type of data used in this research is documentary data. The research data was obtained from the company's financial statements.This researcher obtained data by documenting data from financial reports, then the data was further processed to present the values of the variables studied.So, this research data is not directly obtained from the source.Sources of data with characteristics like this are called secondary data (Sujarweni, 2015;56).This study uses discretionary accruals (DA) calculated using the modified Jones model.Earnings management is measured using discretionary accruals (DACC) which is calculated by excluding total accruals (TACC) and nondiscretionary accruals (NDACC).Discretionary accruals are abnormal accruals that come from management policies.Determination of discretionary accruals as an indicator of earnings management can be described in the calculation stages as follows (Sulistiawan, Januarsi, & Alvia, 2011; 73): 1. Determine the total value of accruals with the formulation : TACCit = NIit -CFOit Information : TACCit = Total accruals of company i in year t NIit= Net profit of company i in year t CFOit= Cash flow from company i operations in year t 2. Determine the parameter values α1, α2, α3 using the Jones model, with the formulation: TACCit = α1 + α2 ∆Revit + α3 PPEit + eit Then, to scale the data, all of these variables are divided by the previous year's assets (Ait-1), so the formula changes to: TACCit/Ait-1 = α1 (1/Ait-1) + α2 (∆Revit/Ait-1) + α3 (PPEit/Ait-1) + eit 3. Calculating NDA values with formulations : In this study, the activeness of the audit committee is expressed in a dummy variable, where code 1 is given if the audit committee meets at least 4 times a year, and code 0 is given if the audit committee holds meetings less than 4 times a year.This variable is measured numerically, which is seen from the nominal number of meetings or meetings held by the audit committee in the current year (Prastiti and Meiranto, 2013).The creation of an effective oversight function of the audit committee is related to the amount of resources owned by the committee.The larger the size of the audit committee, the company will have sufficient resources to monitor the company.The audit committee is a committee formed by the company to carry out the supervisory process regarding the preparation the company's financial statements, which aim to suppress an act of fraud committed by company management.This variable is measured numerically, judging by the nominal number of audit members (Widiastuty, 2016).

Hypothesis development
The frequency of regular meetings can reduce earnings management actions (Pamudji and Trihartati 2010).The audit committee is required to hold meetings 3 to 4 times a year to carry out obligations related to matters relating to the financial reporting process (Mintara 2008).The audit committee that meets less than the minimum number set, has the opportunity to re-occur earnings management (Abbott et al., 2005).

H1:
The number of audit committee meetings has an effect on earnings management.
An audit committee with a sufficient number of members will support the oversight function of the actions taken by agents.The audit committee consists of at least 3 members, both from independent commissioners and outsiders (Financial Services Authority Regulation Number 55/POJK.04/2015).

H2:
The size of the audit committee affects earnings management

Population and Sample
According to Sugiyono (2012; 215) population is a group of people, events or anything that has certain characteristics.The population in this study are all manufacturing companies that publish annual reports and are published on the IDX (Indonesian Stock Exchange).Sampling in this study used a purposive sampling method with predetermined criteria.According to Sugiyono (2012; 2150) Purposive sampling is a sampling technique for data sources with certain considerations, namely the data source is considered to know best about what is expected, making it easier for researchers to explore the object or social situation being studied.The criteria that have been determined by the researcher in selecting the sample are intended so that the sample under study is in accordance with the research objectives.The criteria for this research sample are as follows: 1. Manufacturing companies listed on the Indonesia Stock Exchange during the period studied, namely 2016-2019.2. Regularly publish complete financial reports, as required by researchers.3. Publish financial reports in rupiah.4. Manufacturing companies that generate profits during the period studied.
The type of data used in this study is data derived from documents, namely the financial statements of sample companies.Data is obtained by documenting several items included in the financial statements.The type of data used in this research is documentary data.
The research data was obtained from the company's financial statements.This researcher obtained data by documenting data from financial reports, then the data was further processed to present the values of the variables studied.So, this research data is not directly obtained from the source.Sources of data with characteristics like this are called secondary data (Sujarweni, 2015;56).

Variable Operational Definition and Variable Measurement
The dependent variable or dependent variable is a variable that is affected or becomes the result, because of the independent variable.The dependent variable in this study is earnings management.Lidiawati and Fadjrih (2016) state that detection of the possibility of earnings management in financial reports is examined through the use of accruals.This study uses discretionary accruals (DA) calculated using the modified Jones model.Earnings management is measured using discretionary accruals (DACC) which is calculated by excluding total accruals (TACC) and nondiscretionary accruals (NDACC).Discretionary accruals are abnormal accruals that come from management policies.Determination of discretionary accruals as an indicator of earnings management can be described in the calculation stages as follows (Sulistiawan, Januarsi, & Alvia, 2011; 73): 1. Determine the total value of accruals with the formulation : TACCit = NIit -CFO it Information : TACCit = Total accruals of company i in year t NIit = Net profit of company i in year t CFO it = Cash flow from company i operations in year t 2. Determine the parameter values α1, α2, α3 using the Jones model, with the formulation: TACCit = α1 + α2 ∆Revit + α3 PPEit + eit Then, to scale the data, all of these variables are divided by the previous year's assets (Ait-1), so the formula changes to: TACCit/Ait-1 = α1 (1/Ait-1) + α2 In this study, the activeness of the audit committee is expressed in a dummy variable, where code 1 is given if the audit committee meets at least 4 times a year, and code 0 is given if the audit committee holds meetings less than 4 times a year.This variable is measured numerically, which is seen from the nominal number of meetings or meetings held by the audit committee in the current year (Prastiti and Meiranto, 2013.)The creation of an effective oversight

Research Result
Description of Research Data In this study the population used by the researchers were manufacturing companies listed on the Indonesia Stock Exchange (IDX), data obtained from the official website www.idx.co.id..The method used is purposive sampling method, this method is used to determine the sample according to certain criteria, and companies that do not fit the criteria are not used as research samples.The following table describes the sample research using purposive sampling method.
Descriptive statistics is a field of statistics that studies methods of collecting, compiling and presenting research data.Descriptive statistics are part of the science of statistics that summarizes, presents and describes data in an easy-to-read form so as to provide more complete information.Descriptive statistics only relate to describing or providing information about a data or situation or phenomenon, in other words, only seeing a general description of the data obtained (Ghozali, 2013;19).
In this research, the data analysis technique used is multiple linear regression.This research is included in explanatory research which aims to predict phenomena by testing variables.Before carrying out the regression test, you must first carry out a classical assumption test to test the quality of the data obtained (Ghozali, 2018: 107).
The normality test is a statistical test performed to determine the normality distribution of data in a regression.The regression model must be normally distributed.In this study, data normality testing was carried out using the Kolmogorov-Smirnov test (Ghozali, 2018: 30).Data can be said to be normally distributed if it has a significant level > 0.05.When testing the feasibility of the data, the researcher found 47 outlier data.The data is detected through casewise diagnostics analysis.Therefore, the removal of outlier data was carried out by the researcher.This is to get better data.
The normality test above using the Kolmogorov-Smirnov test, it can be seen that the gain from Asymp.sig (2-tailed) is 0.161 which is greater than the significance level because it is above 0.05.This indicates that the data in this study have been normally distributed and have fulfilled the classical assumption test.The results of the data normality test using the normal probability plot graph are presented in the following figure : Dependent Variable: Earnings Management Using the normal probability plot graph show that the data points spread around the direction of the diagonal line.This shows that the residual data from this study can be considered to be normally distributed and have met the assumptions of data normality.
The multicollinearity test is intended to test whether the data has a high correlation between the independent variables.A regression model can be said to be good or good if there is no correlation between the independent variables.The results of the multicollinearity test can be seen through the Tolerance value and VIF value.Data can be said to be free from multicollinearity problems if the VIF value obtained is below 10 and has a tolerance value greater than 0.10.Therefore, these results can be said that there is no multicollinearity.
The results of the multicollinearity test show that in this study, all of the two variables used had a VIF value below the number 10, namely 1.169.As for the tolerance value in this study above 0.10, namely 0.856.So it can be concluded that the regression model used is free from multicollinearity tests.
This autocorrelation test aims to see the correlation of interfering errors in the t period with the t-1 (previous) period.The regression model can be said to be good if the regression model is free from the autocorrelation test.Autocorrelation testing uses Durbin Watson (DW) with the condition that the regression model does not have autocorrelation if Durbin Watson is greater than the value (du) and less than the value (4-du).Values (du) and (dl) can be through the Durbin Watson table according to the number of K (independent variable) and the number of samples (N).
The autocorrelation test results obtained, the calculated value of Durbin Watson is 1.649.Determining whether or not there is autocorrelation is by comparing the calculated value of Durbin Watson with the value of the Durbin Watson table.Durbin Watson's calculated value is compared with a significance value of 5% of Durbin Watson's value.Where the number of samples found was 113 (n) with 2 (k) independent variables, the result of the du value (upper limit) was 1.7273.This shows that the Durbin Watson value of 1.649 is greater than the value du and less than the value (4 -du) which is equal to 2.2727.Based on the results obtained, namely 1.7273 <2.2727, it can be concluded that the data is free from autocorrelation.

Heteroscedasticity Test
Heteroscedasticity can be defined as a symptom of an unequal variance from one residual observation to another (Ghozali, 2013;139).The results of the heteroscedasticity test above show that the regression model in this study does not indicate symptoms of heteroscedasticity.This shows that the data points spread randomly and do not form a pattern, either above or below the number 0 on the Y axis, the data spreads randomly and does not form a specific pattern.This multiple regression analysis was chosen because it can help predict how much influence the independent variables have on the dependent variable.This study consists of two independent variables, namely the number of meetings and the size of the audit committee.Dependent Variable: Earnings Management Source: SPSS output results, 2021 Based on Table 4.10 the results of the tests that have been carried out, the multiple linear regression equation is as follows: EM = α + ᵦ1Jrit + ᵦ2UKAit + ɛ EM = 0,120 -0,008 Jrit -0,056 UKAit + ɛ Based on the acquisition of the regression coefficient values that have been informed into the regression model formula in this study, the following conclusions are obtained: 1.A constant value of 0.120 indicates that the independent variables are assumed to be constant.2. The first independent variable, namely the number of audit committee meetings, obtains a coefficient value of -0.008.The regression value indicates that when the number of meetings increases by 1 unit, it will decrease the earnings management value by -0.008.3. The second independent variable, namely the size of the audit committee, obtains a coefficient value of -0.056.The coefficient value is significant when the size of the audit committee increases by 1 unit, it will reduce earnings management -0.056..
In this research, the results of simultaneous hypothesis testing prove that together all the independent variables have a significant effect on the dependent variable.To find out which independent variables have a significant influence, a partial hypothesis test is carried out.The following are the results of partial hypothesis testing: From the table it is obtained that the significance values for the two independent variables are 0.072 and 0.001 respectively.By looking at the significance value of each independent variable in the

Figure 4
Figure 4.4.1 Normal P-Plot Graph Based on Figure 4.4.1 above, the results of the data normality test using t

table ,
Comparison of the calculated t value with t table shows results that are consistent with conclusions based on significant values.The results of the comparison of t count with t table can be seen that t count variable Number of Jrit Meetings (X1) gets a value of -1.817> 1.658.It can be concluded that the calculated t value is greater than t table, then H1 is accepted and the UKAit Audit Committee Size variable (X2) obtains a value of -3.474 > -1.817.It can be concluded that the calculated t value is greater than t table.Thus, it can be concluded that and H2 is accepted.This test was carried out aiming to find out whether the independent variable is simultaneously on the dependent variable or not.If the significance value of the results F (sig) > α 0.05 or Fcount > Ftable, it can be concluded that simultaneously there is an influence of the independent variable on the dependent variable.The following are the results of the simultaneous F test that has been carried out, namely: The result of the ANOVA test, significant values are obtained for the variable Number of Audit Committee Meetings (X1) and Size of the Audit Committee (X2) on the Profit Management variable (Y) simultaneously.The significance value of the ANOVA test results is 0.000.This value is smaller than 0.05 or 0.000 <0.05, from the results of the table above it also produces an Fcount with a value of 11.787 and a Ftable of 3.08.It can be concluded that the variables of the number of meetings and the size of the audit committee simultaneously and significantly influence earnings management Looking at the intensity of the independent variables studied in explaining the dependent variable, a coefficient of determination test was carried out.The intensity of the influence can be seen from the Adjusted R Square value.The higher the Adjusted R Square value, the better the regression model used because it indicates the high ability of the independent variable to explain the dependent variable.The result of the coefficient of determination test.In this table, the Adjusted R Square value is 0.162.These results indicate that the variable Number of Audit Committee Meetings (X1) and Audit Committee Size (X2) is able to explain the variable Earnings Management (Y) 16.2%.Meanwhile, the remaining 83% are other factors not observed in the research model.
ConclusionThis study aims to test and obtain empirical evidence regarding the effect of the number of IRJ: