Why People e-file (or Don t e-file) Their Income Taxes

Transcription

1 Why People e-file (or Don t e-file) Their Income Taxes Linda M. Gallant Bentley College lgallant@bentley.edu Mary J. Culnan Bentley College mculnan@bentley.edu Patrick McLoughlin Bentley College mcloug1_patr@ bentley.edu Abstract In 2005, 52% of taxpayers filed their federal returns electronically. This is far short of the IRS goal of having 80% of returns filed electronically by Using the e-commerce technology adoption literature, this exploratory study investigates factors that differentiate e-filers from non-e-filers. People who e- file perceived e-filing to be less useful measured as convenience, and less expensive than non-e-filers, with convenience making the greatest contribution. The results for perceived usefulness were unexpected. People who e-file also perceived e-filing to be safer but less easy-to-use than those who do not e-file; however, these variables were not significant discriminators of e-filers from those who did not e-file. As the costs associated with e-filing differentiate it from most other consumer e-commerce applications, this study contributes to the e-commerce technology adoption literature. Further, the findings suggest that the e-commerce adoption literature is an appropriate theoretical basis for studying adoption of e- government applications. 1. Introduction Governments around the world have developed e- commerce applications to deliver services to citizens and businesses, and to exchange information with other government agencies [1-5]. While many studies of e-government describe these initiatives, little research has investigated whether people actually need, want, or use these services or what factors motivate people to adopt e-government services [4]. One prominent e-government application is the e- filing of income taxes. In the US, federal e-filing began as a pilot program in three metropolitan areas in During the first year, 25,000 returns or 0.02% of all returns filed that year were filed electronically [6]. The IRS Restructuring and Reform Act of 1998 set as a goal for the IRS to have at least 80% of returns filed electronically by In 2005, nearly 52% of tax returns were e-filed with home computer e-filing increasing to 18%. The 2005 Taxpayer Satisfaction Study [7] found that while those who do e-file expressed high levels of satisfaction, the results also suggested a hardening of resistance to e-filing by nonusers. If the goal of 80% e-filing is to be reached, research is needed in order to understand how to promote adoption of e-filing to a wider audience. While practitioner journals in the U.S. have reported on e-filing [see, 8-11] there is an absence of scholarly research on e-filing. This paper is an exploratory study of what factors differentiate people who e-file their income taxes from those who do not e-file. 2. Technology Adoption and e-filing Like other e-commerce applications, e-filing is an IS application where the citizen interacts with a complex IT system. Therefore, we ground our study in the e-commerce technology adoption literature. Recent empirical research on consumer shopping applications found that perceived ease of use (PEOU) and perceived usefulness (PU) explained intended or actual use [12-14]. Based on their findings, these authors also argued that PEOU in particular was taskrelated, with PEOU playing a more important role in information search tasks than purchasing tasks. In one study of e-filing in Taiwan, Wang [15] found that perceived ease of use was a stronger predictor of people s intention to e-file than both perceived credibility and perceived usefulness. This is consistent /07 $ IEEE 1

2 with Wixom and Todd [16] who found that PU was influenced by PEOU. Trust has also been found to influence consumer intentions to adopt e-commerce [13, 14]. A key element of trust is risk: people trust if they act despite the possibility of being disappointed by the actions of another [17]. Given that e-filing is currently voluntary and involves the transfer of highly sensitive personal information, if people perceive e-filing as unsafe or excessively risky, they will be less likely to adopt e- filing over other methods. For example, a 2005 survey conducted by Princeton Survey Research [18] found that due to new concerns about identity theft, a majority of consumers have stopped giving out personal information on the Internet, 30% say they have reduced their overall use of the Internet, and 25% say they have stopped buying things online. Trust is particularly important in transactions where individual lack the ability to monitor or control the actions of the organizations with which they interact. For example, Tan, Pan, and Lim [19] studied the development of an electronic tax filing system in Singapore and found that citizens had restricted knowledge of what happens with their personal data during e-filing in addition to being likely to worry about having sensitive personal data intercepted during the e-filing process. Gefen et. al. [13] argue that PEOU and trust are related as a web interface that is easy to use can convey caring and a commitment to the relationship on the part of the vendor. Further, situational normality, an assessment that a transaction will be a success based on how normal the situation appears to be, also serves to minimize risk. An easy-to-use web interface can promote trust as well as PEOU through situational normality [13]. Finally, e-filing differs from the e-commerce applications of shopping and information search studied in prior research in one important way as e- filers are often charged to e-file. No comparable surcharge exists for Internet shopping or searching, or for use of software in the workplace. As a result, cost is absent from the prior research on both technology adoption and user satisfaction (e.g. [16]). Interviews with e-filing users suggest that cost is also likely to influence the decision to e-file [7]. 3. Research question This study addresses the following research question RQ: What factors differentiate people who e-file from those who do not? The prior theory on e-commerce adoption reviewed above leads us to expect that people who e-file will perceive e-filing to be easier to use, to be more useful, to be less expensive and to pose a lower risk than people who do not e-file. 4. Methodology 4.1 Sample The study is based on a secondary analysis of survey data collected by a tax preparation software company. Participants were selected from the database of customers who previously purchased the company s tax preparation software. Subjects were invited to participate in the survey by . The survey contained a direct link to a web-based survey, which was left active for seven days. For completing the survey, respondents were offered a coupon for a free encyclopedia set. After four days, a follow up was sent to remind people to complete the survey. The study is based on 935 usable responses. The participants were 39% female and 61% male. The remaining demographics were measured using categories. The median age was years, median income was $50,000-59,999; median education was a bachelor s degree; median Internet usage was 5-10 hours per week and median online purchases during the past year was 5-10 purchases. 4.2 Variables The dependent variable, e-filing, was measured by asking people whether or not they had e-filed (yes=38%, no=62%). All of the study participants filed their own taxes versus using a tax practitioner to e-file on their behalf. The independent variables were measured using a series of single-item ascending Likert scales (1=strongly disagree to 5=strongly agree). Means and standard deviations for the independent variables are shown in Table 1. Perceived ease of use was measured as Software for e-filing tax returns is easy to use. Perceived usefulness was measured as E-filing tax returns saves time. The trust variable, risk of e-filing, was measured by two items, E-filing tax returns is [not] safe for my personal financial information. The second item was recoded and the two items were summed to form a scale (r=.51, p<.001). Cost was measured as E-filing tax returns is expensive. 2

3 The study participants who did not e-file were also asked to respond to an open-ended question. These individuals were asked to supply the top three reasons why they did not e-file, where their first response was the most important. 5. Results We used a stepwise discriminant analysis to analyze our data. Discriminant analysis is the appropriate statistical technique for determining if significant differences exist between the profiles of two groups defined by a categorical dependent variable [20]. See Table 1 for the results. The overall discriminant functions are significant (Chi-Square = 45.88, 2 d.f., p. < 001). The function correctly classified 70% of the holdout sample which is greater than the proportional chance criterion plus 25% of 0.65% [20]. Only two of the independent variables were significant discriminators: people who e-file perceive e-filing to be significantly less convenient and less expensive than people who do not e-file. Ease-ofuse and risk were not significant discriminators (See Table 1 for means for the two groups). To assess the significance of the independent variables, the structure correlations were interpreted to assess the relative contribution of each independent variable to the discriminant function, similar to the way factor loadings are used [20]. Convenience makes the greatest contribution, followed by cost. Finally, mean differences were computed for the demographic variables. People who e-file are significantly older, better-educated, report higher income and made more online purchases last year than people who do not e- file. There was no significant difference in hours per week of Internet use between the two groups. An analysis of the open-ended responses provides support for the results of the discriminant analysis. Of 300 total responses, 233 were classified into five categories and these results are shown in Table 2. People indicated that cost was the leading factor deterring them from e-file. Ease of use defined as having problems with using or understanding the process was ranked second. Not being allowed to e- file was ranked third. Other reasons included owing money on their return, needing an extension, or using an accountant. 6. Discussion The findings suggest that the e-commerce technology adoption literature is an appropriate theoretical perspective for studying e-filing. Overall, perceived usefulness measured as convenience, and the cost variables significantly discriminated those who e- filed from those who did not. While people who e-file perceive e-filing to be less expensive, less convenient, less easy to use and safer than those who do not e-file, ease of use and risk were not significant. Interestingly, the results for perceived ease of use and usefulness were in the opposite direction than the theory predicted. What explains these counter-intuitive findings for PU? It may be that e-filing turns out to be more difficult to use than people believed and as result, less convenient and as a result, they modify their beliefs based on their experience. The most recent IRS Taxpayer Satisfaction Study provides support for this conclusion [7]. Less than one-third of those who e- filed in 2005 using an online company or directly using online software volunteered that e-filing was easy or convenient. However, this merits further study. An important contribution of this study is the finding of cost as a significant discriminator of adopters and non-adopters of e-filing. The responses to the open-ended questions in this study and the Taxpayer Satisfaction Study [7] also support the finding for cost as a significant discriminator. As stated previously, cost has not been an issue in prior research on technology adoption. It is an issue for non-workplace applications where usage is voluntary and where cost can serve as a deterrent despite the benefits of usage to the individual or to society atlarge, such as e-filing or the implementation of security software on home computers. As a result, it should be incorporated in future research focusing on technology adoption by consumers in non-workplace settings. 6.1 Limitations This paper is based on a secondary analysis of a survey conducted by a third party and the authors had no control over the survey design. As a result, this exploratory study has a number of limitations. First, the study is based primarily on a convenience sample of one firm s customers, and those who e-filed used a single method. While the results in general are consistent with the results of the Taxpayer Satisfaction Survey [7], a more representative sample might have yielded different results. Second, the variables were measured using single items which lacked appropriate psychometric attributes. Nonetheless, the findings suggest that e-filing is an interesting area for future research with the potential to contribute both to the e- government literature and the e-commerce technology adoption literature. 3

4 6.2 Future Research Future research should replicate this study with a more representative sample and a wider range of variables. In particular, research with better measures of both PEOU and PU tailored to e-filing are needed. For example, prior research (e.g. 21, 14) found that PEOU of the web interface is more important to search tasks while e-filing may more closely resemble a purchase transaction than searching. However, if nonadopters believe e-filing itself is difficult, they may be unwilling to adopt e-filing. Wixom and Todd [16] argue for integrating the user satisfaction literature with the technology adoption literature as technology acceptance provides only limited guidance about how to influence usage through systems design. This would provide a more effective way to assess how system characteristics influence ultimate usage [16]. For example, less than a third of e-filing resisters agree that e-filing is an accurate way to file [7]. Further, there are multiple ways to e-file. People can e-file by using a tax practitioner to prepare their taxes where the practitioner essentially provides chauffeured access to e-filing [22] or e-file their own return. In addition, taxpayers with an Adjusted Gross Income of $50,000 or less can prepare their returns online and e-file for free. As costs vary for each of these methods of e-filing and cost is a major factor in the decision to e-file, future research should include subjects who e-file using a variety of methods as well as people who do not e-file in order to understand the relationship between cost and other adoption factors such as PEOU and PU. CBS News [23] reported that tax returns filed through the free filing program declined 20% from March 2006 compared with the previous year, suggesting that factors other than cost influence the decision to e-file. Future research also needs to further investigate the role of trust in the decision to e-file, particularly given the sensitive nature of tax information and increased concerns about security in the wake of media reports about recent security breaches. This study only measured risk in terms of e-filing being safe. Privacy and security both represent potentially significant risks of e-filing. Future research needs to measure these issues more precisely, particularly since a majority of those resisting even free e-filing cite concerns about the privacy and security of e-filing [7]. In conclusion, the IRS faces a major challenge in reaching the goal of having 80% of tax returns e-filed by 2007 given only 52% of taxpayers filed online in 2005 [6]. While the most recent Taxpayer Satisfaction Survey (IRS 2005) indicates that a majority of e-filers continue to report high levels of satisfaction, substantial numbers also report that e-filing could be easier to use and less expensive. Further, the IRS data suggest a hardening of resistance to e-file among the remaining non-users. Only 26% of lapsed users and 22% of people who have never e-filed agree completely that e-filing is a better way to file their federal income taxes. Accuracy, trust (privacy and security) and ease of use are lower for resisters than for taxpayers who do e-file. Additional empirical research on e-filing grounded in the technology adoption literature can not only help the IRS increase the level of e-filing, but should also contribute to our understanding of why consumers adopt e-commerce applications more generally. 7. References [1] Davison, R. M., Wagner, C., & Ma, L. C. K. (2005). From government to e-government: A transitional model. Information, Technology, & People, 18(3), [2] Grant, G. & Chau, D. (2005). Developing a generic framework for e-government. Journal of Global Information Management, 13(1), [3] Jaeger, P. T. (2003). The endless wire: E- government as global phenomenon. Government Information Quarterly, 20, [4] Saxena, K. B. C. (2005). Towards excellent in e- governance. International Journal of Public Sector Management, 18(6), [5] Stoltzfus, K. (2005). Student research papers: Motivations for implementing e-government: an investigation of the global phenomenon. Proceedings of the 2005 national conference on Digital government research, [6] Weigelt, M. (2006), E-File goal too ambitious. FCW.COM, February 27, available at: [Accessed June 14, 2006]. [7] Internal Revenue Service (2005), Findings from the 2005 Taxpayer Satisfaction Study, Publication 4241 (7/05), available at: [Accessed June 13, 2006]. 4

5 [8] Culligan, M., Scutellaro, J. F., & Szargowicz, L. A. (2002, November). CPAs can no longer postpone signing up for the IRS electronic fling program: It s time to e-file. Journal of Accountancy, [9] Lombardo, C. (2001, November). Business bundling: Tax vendors are ready for business tax to move on line. Accounting Technology, [10] Needleman, T. (2004, October). The e-filing experience: E-Filing has only gotten better with advantages for all. Practical Accountant, [11] Sponge, L. & Trewin, J. (2003, December). The basic of E-File. Strategic Finance, 1-2. [12] Gefen, D. & Straub, D. W. (2004). Consumer trust in B2C e-commerce and the importance of social presence: Experiments in e-products and e-services. Omega: The International Journal of Management Science, 32, [13] Gefen, D., Karahanna, E., and Straub, D.W., (2003). Trust and TAM in online shopping: An integrated model, MIS Quarterly, 27(1), [14] Pavlou, P. A. & Fygenson, M. (2006), Understanding and predicting electronic commerce adoption: An extension of the theory of planned behavior, MIS Quarterly, 30(1), [15] Wang, Y. S. (2002). The adoption of electronic tax filing systems: An empirical study. Government Information Quarterly, 20, [18] Princeton Survey Research (2005). Leap of Faith: Using the Internet Despite the Dangers. Available at: [Accessed June 13, 2006]. [19] Tan, C. W., Pan, S. L., & Lim, E. T. K. (2005). Managing stakeholder interests in e-government implementation: Lessons learned from a Singapore e- government Project. Journal of Global Information Management, 13(1), [20] Hair, J. F., Anderson, R.E., & Tatham, R.L. (1987), Multivariate Data Analysis with Readings, 2 nd Edition. New York: Macmillan Publishing Company. [21] Gefen D. and Straub, D.W. (2000). The relative importance of perceived ease-of-use in IS adoption: A study of e-commerce adoption. Journal of the Association for Information Systems, 1(8), [22] Culnan, Mary J. (1983). Chauffeured versus end-user access to commercial databases: the effects of task and individual differences. MIS Quarterly, 7(1), [23] CBS News (March 2006). Free e-filing free falling: IRS filing of electronic returns through program is down 20 percent. Available at: n shtml [Accessed June 13, 2006]. [16] Wixom, B. H., & Todd, P. A. (2005). A theoretical integration of user satisfaction and technology acceptance. Information Systems Research, 16(2), [17] Mayer, R. C., Davis, J.H. & Schoorman, F.D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3),

6 Table 1 Discriminators of e-filing (Stepwise Analysis) Eigenvalue Canonical Wilks Chi- Squared df Sig. Correlation Lambda Group: 0 (Not efile) 1 (efile) Centroids: Independent Variables* Mean s.d. Mean efile=0 Mean efile=1 Standardized Canonical Coefficients Structure Matrix: Pooled-within- Groups- Correlations Convenience Expensive Ease of Use Risk (Safe) *All 5-item ascending scales: 1=Strongly Disagree to 5=Strongly Agree. Risk measured using a two-item scale. Table 2 Total Responses to Why Respondents Did Not E-file in the Previous Tax Year Rank Category Number of Responses Percentage 1 Cost % 2 Problems using/understanding % 3 Not Allowed to E-file 81 20% 4 Owed Money 56 14% 5 Extension Needed 18 4% 6 Used Accountant/Face-to-Face 8 2% Total % 6