Tax Administration: IRS Improved Performance in the 2004 Filing Season, But Better Data on the Quality of Some Services Are Needed

Both the Internal Revenue Service Commissioner as well as the Commissioner of the Wage and Investment Division have acknowledged that data obtained by direct management observation may be inherently biased because of employee's knowledge that they are being observed.

IRS has taken actions to address these delays and further implement contact recording to collect data on the quality of services at walk-in sites. According to IRS officials, since its pilot of contact recording in 2005, IRS has continued deploying contact recording as a method of collecting data to measure accuracy of account and tax law assistance. We recently subsequently reported that while IRS appears to be on schedule based on its implementation plan for contact recording, it has previously experienced delays implementing other parts of its quality review program. As of July 2007, IRS has deployed contact recording to 127 of its 401 walk-in sites; these sites account for 46 percent of IRS walk-in site contacts. IRS will continue to install contact recording at 175 additional walk-in sites in fiscal year 2008 and have contact recording fully deployed by the end of fiscal year 2009. With contact recording, IRS will have better data collection and analytics of account and other assistance. As a result, IRS and the Congress have better information and understanding of the quality of service IRS is providing at walk-in sites.

GAO was primarily concerned with delays in developing a means of monitoring the quality of return preparation services at volunteer sites. In time for the 2005 tax filing season, IRS had two methods in place to assess quality at volunteer sites: 1) firsthand observational reviews conducted by IRS representatives of volunteer/taxpayer interaction at a sample of sites (including a review of the taxpayer's return and paperwork for accuracy), 2) "mystery shopping" conducted by representatives from partner groups posing as taxpayers seeking return assistance. Notably, however, IRS implemented neither of these methods as planned for the duration of the 2005 tax filing season. First, IRS abandoned observational reviews by the second week in February because of concerns raised by the National Taxpayer Advocate and other partners. Second, IRS conducted only 14 of the planned 100 mystery shopping reviews, which did not yield a statistically valid sample.

In IRS's response to the report, the Commissioner noted that during the 2004 filing season, the Electronic Tax Law Assistance (ETLA) option was inadvertently placed in a prominent Web site location, causing the Service to experience an unanticipated demand that exceeded its ability to deliver the service levels it intended. The Commissioner added "We fully recognize the relationship between the prominence of the ETLA web location and the need for commensurate staffing." Review of the IRS website in 2006 shows that the ETLA feature is not prominently placed at this time, consistent with the Commissioner's remarks that the ETLA option was designed to serve a very limited number of customers. Furthermore, IRS's performance in responding to tax questions received via Internet during the 2005 and 2006 filing season was better than its performance during the 2004 season, suggests that staffing levels are appropriate given current web placement of the feature. Specifically, during the 2005 filing season IRS took an average of only 1.2 business days to respond to tax law questions submitted via the Internet. During the 2006 filing season, IRS took an average of 2.2 days to respond. On the other hand, in the 2004 filing season, IRS took, on average, over three days to respond to tax law questions submitted via Internet.