E-Filing vs Non E-Filing: Taxpayers’ Perceptions of Use- fulness, Ease, Security, Confidentiality, and Satisfaction

Objective – Tax reporting through the Annual Tax Return (SPT) is one of the obligations of taxpayers as mandated by the Indonesian Taxation Law. Along with the development of science and technology, technological innovation has been adopted as a form of internet-based taxation service called the e-filing system. This study is aimed to compare the perceived usefulness, convenience, security and satisfaction of taxpayers between users of e-filing and non e-filing. Design/methodology – Primary and secondary data are collected through survey method in Tax Office (KPP) Pratama Purwokerto, Central Java Province, Indonesia. The registered taxpayers using e-filing are 17,831 and the non e-filing are 44,332. Paired t-test was used to compare each variable between users of e-filing and non e-filing. Results – The findings reveal that there is no difference in perceived usefulness between the users of e-filing and non e-filing in that reporting tax return online and manual are equally perceived useful. Nevertheless there is a difference in the perceived convenience and security whereby users perceived that the use of e-filing is more convenient, secure and confidential than the manual. Furthermore there is no difference in the perceived satisfaction whereby users are equally satisfied either with the e-filing and non e-filing. As a whole, there are differences of perception between users of e-filing and non e-filing, hence it can be concluded that the benefits of the usefulness of e-filing is influenced by variables of usability, convenience, security, confidentiality and satisfaction of taxpayers on e-filing.


Introduction
One of the leading national incomes in supporting the development of a country is tax. Tax is a vital income in a country, without tax revenue the state will not run well (Aditya, 2018). In Indonesia, based on the Revised Draft of State Revenue and Expenditure Budget (RAPBNP) 2015, domestic revenue is targeted at IDR 1,765,662.2 billion. The domestic income consisted of tax revenues of IDR 1,484,589.3 billion and PNBP of IDR 281,072.9 billion. Tax revenue accounts for 84% of total domestic revenue in Indonesia, the largest source proportion of state revenue. Every citizen as a taxpayer, has responsibility of taxation as a reflection of state obligations. This is in accordance with the self-assessment system adopted in the Indonesian tax system (Damayanti, T, Subekti, & Baridwan, 2015). This collection system makes every taxpayer have obligation to notify amount of their owed tax to the Directorate General of Taxes (DGT) through tax returns (SPT) before pay it.
Reporting the amount of tax through SPT is one of the obligations of Taxpayers, this mandated in Indonesian Taxation Law No. 6 of 1983 as last amended by Law no. 28 of 2007 in article (3). Tax returns must be submitted by the taxpayers to DGT through the Tax Service Office (KPP) manually beforehand. However, along with the development of science and technology especially in computerization and internet terms, DGT has to adopt a new technological innovation (Hastuti, Suryaningrum, Susilowati, & Muchtolifah, 2014). One form of internet-based tax service that currently being implemented in Indonesia is the e-filing system. It is a service system for submitting Monthly SPT Annual SPT in electronic forms via device technology (Sari, 2013).

JAROE VOL. 3(2)
The submission of SPT through e-filing basically is one of DGT's effort to provide easy services for taxpayers to report their tax. However, does not mean every taxpayer are willing to move from a manual system to an e-filling facility. One of the determinants of the system acceptance is the user satisfaction of it. User satisfication is also important to know level of system acceptance. Being satisfied of using e-filing, the taxpayers will be interested in reusing it again. Likewise, if taxpayers feel disappointed after using e-filing, they will become less interested in using it (Noviandini, 2012).
The CNN Indonesia website in 2014 stated that the number of individual taxpayers was 16.97 million (Jati, 2015). A total of 8.68 million individual taxpayers from the total have an obligation to submit SPT. The convenience provided by the e-filling system was implemented in the previous year has succeeded in attract individual taxpayers from conventional reporting convert to e-filling. The evidence is DGT's data recorded a decrease of conventional obligatory report method from 6.51 million SPT in the previous year into 6.21 million SPT (Jati, 2015).
At the deadline for individual taxpayers annual SPT reporting in March 2015, the number of individual taxpayers reporting their taxes using the e-filling system at KPP Pratama Purwokerto reached 6,758 taxpayers (Abdul, 2015). It exceeded the target estimated by KPP Pratama Purwokerto. However, it is not yet 100 percent of the total number of taxpayers who already had e-filling accounts. It had not reached half of the total of 14,000 taxpayers who had e-filling accounts.
Based on the problems and research that has been described, this study is interested in analyze individual taxpayers perceptions of the use of e-filing and non-e-filing systems for annual taxes. It refers to the research conducted by Desmayanti (2012) on the factors influencing the use of e-filing facilities by taxpayers as a means of submitting monthly online and realtime SPT.

Theory of Planned Behavior
Theory of Planned Behavior (TPB) developed by Ajzen in 1980 is a development of Theory of Reasoned Action (TRA) in 1967. TPB strengthens TRA by adding a new construct called perceived behavioral control. TPB shows that human action is directed by three kinds of beliefs consisting of behavioral, normative, and control beliefs. Overall, they will generate interest in individual behavior which will further determine the individuals in using the system (behavior). This theory can explain the basic perception of the perceived ease with which individual taxpayers will lead to the decision to use e-filing on the basis of the ease of operation.

Task-Technology Fit (TTF)
TTF was developed by Goodhue and Thompson in 1995. Its suitability theory is defined as an ideal profile that formed from a set of task dependencies that are internally consistent with the technology elements used that result in task performance (Jogiyanto, 2007). It directly considers technology to have a positive impact on individual performance, it also be used for calculate technology's capabilities match the tasks generated by users (Desmayanti, 2012). It can be used as a reference point that the level of security and confidentiality is one of the positive benefits that the e-filing system can provide to every individual taxpayer who uses e-filing.

Technology Acceptance Model (TAM)
Davis introduced the Technology Acceptance Model (TAM) in 1989. It is a theory that explains use of information technology systems is considered very influential. It also explains the individual acceptance of the use of information technology systems. It is a model used to predict user acceptance of technology based on two variables that consist of perceived usefulness and perceived ease of use. The response of information technology users is important, this because of acceptance or rejection of information technology must be ascertained before its success rate known. User responses are very close to a variety of user problems and potential rewards received from the implementation of information technology in user activities about taxation (Desmayanti, 2012).
TAM describes two factors that predominantly influence technology integration. The first factor is the perception of usefulness and the second factor is the perception of ease of use of technology. It is used in this study as a basis for taking variables that consist of perceived usefulness and perceived ease of use affect individual taxpayers attitudes behavior in the use of information technology. This research also will determine behavior of these individuals to use the information technology. Davis (1989) in Jogiyanto (2007) defined perceived usefulness as the extent to which a person believes that using a technology will improve his job performance. When it is related to TAM, if someone feels that the use of technology will improve their performance, he will continue to use it. Devina & Waluyo's research (2016) explains that the e-Filing system can help taxpayers cut costs and save time because it can be done anytime and anywhere as long as they are connected to the internet. Similar research results were also presented by Desmayanti (2012), Dyanrosi (2015), Lie & Sadjiarto (2013), Wibisono & Toly (2014) which stated that perceived usefulness had a significant effect on the intensity of e-filing usage behavior. H1 : There is a difference in perceived usefulness between e-filing and non-e-filing users in reporting individual annual taxes.

Perceived Ease of Use
Perceived ease of use in TAM is defined as the extent to which a person believes that using a technology will be free of effort (Davis, 1989). It is also a belief about decision making. If someone believes that information systems are easy to use, he will use them, and vice versa. Devina & Waluyo (2016) explains that the e-Filing system is able to make taxpayers not have to queue at the location for sending SPT data. SPT reporting through e-Filing can be sent directly to the DGT's database as long as they are connected to the internet. It can make taxpayers easier to carry out their tax obligations. In line with these results, research conducted by Desmayanti (2012), Dyanrosi (2015), Lie & Sadjiarto (2013) and Wibisono & Toly (2014) shows that perceived ease has an effect on the behavior of taxpayers in using e-filing. H2 : There is a difference in the perceived ease between e-filing and non-e-filing users in reporting the annual personal tax.

Security and Privacy
Hamlet & Strube (2000) state that security means that the use of information systems is safe would make risk of losing data or information is very small and then risk of theft steal the data is low. Meanwhile, confidentiality means that all matters relating to the user's confidentiality of personal information is guaranteed. An information system is said to be good if its security is reliable. It is measured through user data that is safely stored by an information system. The confidentiality of user data must be maintained by means of data stored by the information system so that other parties cannot access user data freely (Dewi & Khomalyana, 2009). The TTF theory explains that there is an impact in helping individuals do tasks directly. The principle in this theory is that technology has a positive impact on individual performance and it can be used if the technology's capabilities match the tasks generated by users (Desmayanti, 2012). The majority of users is not equipped with a thorough understanding of the security and confidentiality risks of e-filing. Users assume that the ASP has paid attention to their security and confidentiality, even though they do not know how strong the technology tools are for the security and confidentiality of SI from e-filing JAROE VOL. 3(2) (Sugihanti, 2011). In this case Wibisono & Toly (2014) explained that the security and confidentiality of user data can affect taxpayers interest in using e-filing. H3 : There are differences in security and confidentiality between e-filing and non-efiling users in reporting individual annual taxes.

Tax Payer Satisfaction
According to Seddon & Kiew (1996) in Noviandini (2012) satisfaction is a clean feeling of being happy or unhappy in receiving an information system from the overall benefits that someone expects. It is generated from interactions with information systems. Provide useful and quality information (Tajuddin, Hassan, Othman, & Razak, 2020), when users are satisfied with the system provided, they will tend to have loyalty to use it (Auraningtyas, 2012). The measure of user satisfaction in information systems can be reflected in the quality of the system (Guimaraes, Igbaria, & Lu, 1992;Yoon, Guimaraes, & O'Neal, 1995). If the quality of the information system is good based on the perceptions of the user, they will tend to feel satisfied in using it. Research by Istianingsih & Utami (2009) provides empirical evidence that the quality of information systems has a positive and significant effect on user satisfaction. It is predicted that the higher the quality of the information system used, the higher the level of satisfaction of the information system end users will be. Seddon & Kiew (1996) explained that the level of information quality is in line with the level of end-user satisfaction of the information system. User satisfaction will be affected by information systems that are able to produce information that is timely, accurate, as needed, relevant and fulfills the quality of information criteria. It is in line with the results of research conducted by Noviandini (2012) that there is a positive influence on taxpayer satisfaction in using e-filing. H4 : There is a difference in satisfaction between e-filing and non e-filing users in reporting the annual personal tax.

Research Method
It is a hypothesis testing research using a survey method, where information is collected from respondents using a questionnaire (Singarimbun & Effendi, 2008). Meanwhile, the object of this research consists of perceptions of usefulness, convenience, security and confidentiality, and satisfaction of taxpayers with e-filing and non-  The types of data in this study are primary and secondary data. Data were collected using a survey method from several respondents who have registered at KPP Pratama Purwokerto. The questionnaire is measured using a Likert scale to measure perceptions of the usability, convenience, security and confidentiality as well as taxpayer satisfaction.
The population in this study were 17,831 taxpayers of e-filing users and 44,331 of non-e-filing users, a total of 62,162 registered taxpayers at KPP Pratama Purwokerto. The sampling technique was purposive sampling technique, namely taking samples from the population based on certain criteria.

Analysis Technique Validity test
Validity testing is applied to determine the strength of each question item in the distributed questionnaire that can reveal the variables under the study, then analysis of validity of the question is carried out. Validity testing was carried out using the Karl Pearson Product Moment method Husein (2003) with a level of significance α=0.05 degrees.

Reliability Test
It is a value that shows the consistency of a measuring device in measuring the same symptoms. The technique used is Cronbach Alpha with the level of significance α = 0.05.

Descriptive Statistical Method
In this case, the data analysis is carried out to the data that has been collected by conducting descriptive statistics through descriptions of the research data seen from the mean (mean), standard deviation, variance, maximum, minimum (Ghozali, 2005).

T-test analysis
In orfer to test differences in perceptions of usefulness, convenience, security and confidentiality and satisfaction of taxpayers of e-filing and non-e-filing users, the ttest is used as follows (Sugiyono, 2008).

Research Overview
There are 17,831 taxpayers who are e-filing users who have registered at KPP Pratama Purwokerto. Meanwhile, there are 44,331 taxpayers who still report their SPT manually. The pilot test was conducted before the questionnaire was given to respondents to test the feasibility of the research instrument. The number of pilot test respondents in this study were 30 taxpayers from both e-filing and non-e-filing users.
The questionnaires in this study were distributed from 25 June 2015 to 5 August 2015. It was distributed directly to e-filing and non-e-filing users respondents. One hundred copies of questionnaires were distributed to the two sample categories. The number of copies provided for the sample categories of e-filing and non-e-filing users is 50 copies each. Based on the data in table 1, it is seen that there were 5 copies of the questionnaire that were not returned and there were 3 copies of the incomplete questionnaire. Therefore, the total number of questionnaires distributed in this study has a response rate of 92%. Based on the rate of return, then it is divided equally between e-filing users (46 copies) and non-e-filing users (46 copies).

JAROE VOL. 3(2)
Based on the questionnaire that were returned and processed, personal data of respondents were obtained. The number of male e-filing users was 18 people or 39.13% and female respondents were 28 people or 60.87%, while the number of male non-efiling respondents was as many as 24 people or 52.17% and 22 people were female respondents of non-e-filing or 47.83%. Meanwhile, most of the users of e-filing and none-filing have undergraduate degrees, namely 26 people or 56.52% for e-filing users and 23 or 50% of non-e-filing users. Fifteen diplomas or 32.61% were e-filing users and 13 people or 28.26% were non-e-filing users. Five people or 10.87% of e-filing users were respondents at the high school education level (SMA) and 10 people or 21.74% of none-filing respondents.

Descriptive Analysis
From the number of questionnaires collected, the respondents' responses to the variables studied can be illustrated as follows: Perceived Usability Variable (X1) No e-filing and non e-filing users claim that tax reporting by e-filing is useless. Most of the respondents answered that both conventional personal tax reporting and e-filing are useful. It shows that respondents strongly agree with a series of statements based on indicators of perceived usefulness in e-filing to increase effectiveness, and to make transactions shorter. It shows that the perceived usefulness in the attitude of using efiling is already in the good category.

Perceived Convenience Variable (X2)
Most of the respondents answered 'easy' to statements on the perceived convenience variable. It shows that the respondents strongly agree with the series of statements based on the perceived ease of use indicators that e-filing is easy to apply, for example making transactions easier and time efficient. It shows that the perceived ease of use of e-filing is in the good category.

Variable of Security and Confidentiality Perception (X3)
Most respondents answered strongly agree with the statement on the variable perception of security and confidentiality. It shows that the respondents strongly agree with the series of statements based on the security perception indicator, namely that efiling can be more accurate in making tax payments.

Perceived Satisfaction Variable (X4)
Most respondents answered strongly agree with the statement on the security perception variable. It shows that respondents strongly agree with a series of statements based on indicators of perception of satisfaction, namely e-filing has an added value in

Validity Test
The pilot test was conducted to calculate the value of validity and reliability for each variable and to remove invalid questionnaire items before the research was carried out. After the validity and reliability values of each question item are known, the items in the questionnaire that have met the validity and reliability requirements are again distributed to the sample of research respondents. Table 2, table 3, table 4, table 5 are the results of the validity test on question items that have passed the validity test on the pilot test prior to conducting the study. Table 2 is the result of the validity test of data processed from questionnaires distributed to respondents. The results of the validity test show that rcount value of the product moment correlation of all questionnaire questions on the perceived usefulness variable of e-filing and non-e-filing user respondents has a coefficient greater than the critical value (rtable) of 0.361 at the 95% confidence level. Therefore, all question items are said to be valid and can be used as a data collection tool.

Items rcount (e-filing users) rcount (non e-filing users) rtable Explanation
Question 1 Table 3 shows that the rcount value of product moment correlation of the questionnaire question item on the perceived convenience variable with e-filing and non-e-filing user respondents has a coefficient greater than the critical value (rtable) of 0.361 at the 95% confidence level. Thus, the question items are said to be valid and can be used as a data collection tool.     Table 4 shows that the rcount value of the product moment correlation of the questionnaire question items on the variable security and confidentiality perceptions with e-filing and non-e-filing user respondents has a coefficient greater than the critical value (rtable) of 0.361 at the 95% confidence level. Thus, the question item can be said to be valid and can be used as a data collection tool.
Items rcount (e-filing users) rcount (non e-filing users) The summary in table 5 shows that rcount value of the product moment correlation of the questionnaire questions on the variable perception of satisfaction with e-filing and non-e-filing user respondents has a coefficient greater than the critical value (rtable) of 0.361 at the 95% confidence level. Thus, the question items are said to be valid and can be used as a data collection tool.

Explanation
Perception  Table 6 shows that the Cronbach Alpha coefficient value of the e-filing and non-efiling user groups for the perception of usability variable, the perception of convenience variable, the perception of security and confidentiality variable and the perception of taxpayer satisfaction variable are respectively greater than 0.600. Therefore, all questions for each variable can be declared reliable and can be used as a data collection tool.

159
The first hypothesis states that there is no difference in perceived usefulness between individual taxpayers' e-filing and non-e-filing users in reporting their annual tax returns. The results of the hypothesis test of the perceived usefulness variable in table 7 show the significance value of the difference in perception of p (0.081583)> 0.05. Thus, it can be concluded that there is no difference between e-filing and non-e-filing users for the perceived usefulness variable. Therefore, H1 is rejected. It shows that both e-filing and non-e-filing user respondents stated that both online and manual SPT reporting are useful.
The second hypothesis states that there is a difference in the perceived ease of efiling and non-e-filing individual taxpayers in reporting their annual tax return. The results of the hypothesis test for the perceived ease of perception variable show a significant difference in perception of p (0.045911) <0.05. Thus, it can be concluded that there are differences between e-filing and non-e-filing users in the perceived convenience variable. Therefore, H2 is accepted.
The third hypothesis states that there are differences in perceptions of security and confidentiality between individual taxpayers of e-filing and non-e-filing in reporting their annual tax returns. The results of the perception test in appendix 12 show the significance value of the difference in perception of p (0.019801) <0.05 so that for the variable of perceptions of security and confidentiality between e-filing and non-e-filing users, it can be concluded that there are differences between the two. Therefore, H3 is accepted.
The fourth hypothesis states that there is no difference in the perception of satisfaction between e-filing and non-e-filing users of individual taxpayers in reporting their annual tax returns. The results of the perception test in appendix 13 show the significance value of the difference in perception of p (0.977376)> 0.05. Thus, for the variable of perception of satisfaction between e-filing and non-e-filing users it can be concluded that there is no difference between the two. Therefore, H4 is rejected.  Table 8 shows the significance value of the difference in perception of p (0.016) <0.05. Therefore, it can be concluded that there are differences in perceptions between respondents using e-filing and non-e-filing users. This test is an overall test of the four dimensions of perception.

Research Discussion
Based on the result of the hypothesis test, the first hypothesis is rejected. It shows that there is no difference in perceived usefulness between e-filing and non-e-filing users in reporting their annual tax returns. It is because taxpayers believe that both using e-filing and manually in reporting their SPT are useful and are able to improve performance. The results of this study are certainly not in line with research conducted by Desmayanti (2012), Devina & Waluyo (2016), Dyanrosi (2015), Lie & Sadjiarto (2013), Wibisono & Toly (2014), which explain that perceived usefulness has a significant effect on the intensity of the behavior of using e-filing.
Furthermore, the result of the second hypothesis test shows that there are differences in the perception of ease between e-filing and non-e-filing users of taxpayers. From the results of different perceptions, it is known that the use of e-filing can increase time and cost efficiency compared to manual tax reporting. Another benefit is the relatively shorter time required for verification of the submitted tax reporting files compared to conventional ones. However, the e-filing system does not provide more transaction effectiveness than conventional one. The weakness that most taxpayers feel about this system is the absence of direct communication when the taxpayer needs ver-  (2) ification. In the future, of course, this is the reason to provide live chat facilities compared to only the contact us menu. When taxpayers have difficulties, they can be guided directly visually on the online site with live chat. The ease of submitting online tax returns is proven by the fact that taxpayers feel the ease in submitting tax returns obtained from e-filing. It is in line with the research of Devina & Waluyo (2016), Desmayanti (2012), Dyanrosi (2015) Lie & Sadjiarto (2013), Wibisono & Toly (2014) showing that perceived ease of use has a positive and significant effect on attitudes to technology use.
The result of the third hypothesis test in this study indicates that there are differences in perceptions of security and confidentiality between e-filing and non-e-filing user taxpayers. It proves that the taxpayers feel credibility in using the e-filing system. The results of this study are in accordance with the findings stated by Istiarni (2014) which states that there is a belief that a system can help with everything he needs and can keep things that are considered important for them. TTF theory explains how technology has an impact in helping individuals do tasks. Regarding the belief in the e-filling system that can help in carrying out tax obligations, taxpayers feel that the e-filling system is reliable in providing security and comfort.
If the taxpayers feel that it can be fulfilled, of course, a sense of satisfaction will arise along with its use. TTF theory explains how technology has an impact in helping individuals do tasks. Directly it holds that technology has a positive impact on individual performance. The result of the fourth hypothesis test is that there is no difference in the perception of satisfaction between e-filing and non-e-filing users. It is because taxpayers are served well both in online and manual systems, starting from the quality of service, system and information they get. Therefore, they are equally satisfied in reporting their SPT.

Conclusions, Recommendations, and Limitations
Conclusions can be drawn based on the results of research and discussion. First, there is no difference in perceived usefulness for taxpayers who use e-filing and non-efiling applications in reporting SPT. Second, there are differences in the perception of the ease with which taxpayers using e-filing and non-e-filing applications in reporting SPT. It means that perceived ease of use affects the level of e-filing usage. The higher the level of perceived ease of use, the more frequent taxpayers will use e-filing. Third, there are different perceptions of security and confidentiality for taxpayers who use efiling and non-e-filing applications when reporting SPT. Fourth, there is no difference in the perception of satisfaction in using e-filing and non-e-filing applications by taxpayers in reporting SPT. It means that e-filing and non-e-filing users are equally satisfied in reporting SPT both online and manually.
E-filing and non-e-filing user taxpayers have the same perception about online and manual facilities. Therefore, KPP Pratama Purwokerto still has to provide the same service for both e-filing and non-e-filing users. Features need to be added to review the annual tax return that has been sent as a comparison if taxpayers want to make corrections to the annual tax return or fill in the annual tax return in the following year, and to store permanent data, such as family members or assets so that taxpayers do not need to refill the data each time filling out the SPT. Another feature to add is the loading indicator on the e-filing application during the transitioning process from one step to another. It is important especially for taxpayers who have a limited internet connection to avoid the impression that the application suddenly hangs while being loaded. The security and confidentiality factor of using e-filing makes it a little difficult when taxpayers are required to fill in a security code or captcha without being given the opportunity to ask for an advanced code. The DGT should continue to disseminate the e-filing application so that taxpayers who have not used e-filing will use it. This certainly can make the DGT and the taxpayers themselves easier to carry out their obligations. This research contains several limitations when being conducted, such as the existence of Article 34 of Law Number 6 of 1983 concerning general provisions and taxation procedures as amended several times, most recently by Law Number 16 of 2009, which explains that taxpayer data is protected by law and officials are prohibited from informing taxpayer's data to other parties so that the sample used is limited. Furthermore, respondents consider that all matters relating to taxes are strictly confidential, so that the sample obtained is relatively small, even though it meets the minimum required amount. Future research is expected to add other variables, such as experience and readiness of information technology that can clarify research results. In addition, researchers are expected to be more convincing taxpayers of the confidentiality of the data provided.