Detailed Information on the
Internal Revenue Service Criminal Investigations Assessment

Implement the Investigative Scanning and Document Management System (ISDM). Fall 2008 Update: Due to technical issues, such as performance and movement of files, the pilot for the Atlanta and San Antonio filed offices was delayed until January 2009. The technical issues were either resolved or significant progress was achieved. The pilot will conclude in March 2009 and production rollout will commence in either May or June of 2009. The complete rollout to all field offices is still expected by December 2010.

Implement a pilot program to improve the effectiveness of the Suspicious Activity Report (SAR) Review Teams. Fall 2008 Update: The Task Forces are in various stages of operation, some having??secured??dedicated space and task force officers and others that continue to negotiate space and participation issues.?? Funding was secured for FY 2009 to continue funding the need of the established task forces complete and assist those that are still in the formative stage.?? A review of the initiative will take place after the second quarter of FY 2009 with the goal of expanding the program to additional locations in FY 2010.

New Improvement Action: Criminal Investigative Management Information System (CIMIS) version 2.5 is scheduled for release in April 2009. This version will correct identified defects across CIMIS and incorporate suggestions from the Field including: Modifications to: the equipment activity, including equipment accountability, equipment reports and consolidation of the equipment notifications; the Special Agent in Charge/ Assistant Special Agent in Charge operational review activity and the National Query activity. Other corrections include fixing various defects in the Investigation/Correct Status activity, to Investigation Report INV017 and with the User Defined Reports Activity. This release also includes SSN masking - masking of the employee (or non-employee SSN) and changes requested by the Asset Forfeiture Tracking and Retrieval (AFTRAK) team necessary for the AFTRAK/CIMIS future integration.

Improve program performance by exploring methods for measuring the impact of criminal investigations on tax compliance. IRS will report on its progress by the end of 2006.? The study used sentences, fines, and probation related to CI case convictions. The number of convictions is a current measure. IRS-Criminal Investigation will continue to use the number of convictions as an outcome measure because it was validated by the study as having a strong correlation and positive impact on voluntary compliance. The study result does not affect the target number of convictions.

Improve program performance by implementing a new information management system in 2006 to enhance investigative case tracking and improve efficiency.

Improve program performance by developing methods to improve case prioritization in 2006 to ensure that cases yield the greatest impact on compliance.

Criminal Investigation Management Information System enhancements to address the consolidation of Areas, offices, and the movement of workgroups (CIMIS 2.2) was completed and implemented in April 2008.

Implement upgrades to the infrastructure and components of the CIMIS to keep the technology current and prepare the system for integration with other planned CI systems and CIMIS system upgrades. In addition, implement modification of the application code to allow other CI trusted applications to use the code and data.Upgrades are schedule for release in November 2008. Fall 2008 Update: CIMIS version 2.4 was successfully deployed to production on September 28, 2008.

Measure: Number of convictions

Explanation:This is a count of the number of criminal convictions resulting from CI investigations. This measure is used as a proxy for CI's impact on compliance, its ultimate outcome goal. CI's impact on compliance cannot be directly measured, but academic research suggests that higher levels of convictions/criminal sentences lead to higher compliance.

Measure: Conviction rate

Explanation:Conviction rate is defined as the number of convictions divided by the number of indictments. A high conviction rate is an indicator of the quality of CI's investigative work.

Measure: Conviction efficiency rate

Explanation:Conviction efficiency rate is the cost of CI's program divided the number of convictions.

Measure: Voluntary Compliance Rate (individual income taxes)

Explanation:Percent of individual income taxes due that are reported and paid on time without IRS enforcement action. Data is based on periodic audits of a random sample of tax returns.

Measure: Percent of Taxpayers stating that it is OK to cheat on their taxes

Explanation:A random sample of taxpayers is annually surveyed and asked "How much, if any, do you think is an acceptable amount to cheat on your income taxes?" This measure is used as a proxy for voluntary compliance. It reports the percentage of taxpayers answering either "as much as possible" or "a little here and there." The margin for error is plus or minus 3 percent. 2005 data is not yet available.

Is the program purpose clear?

Explanation: Internal Revenue Service (IRS) Criminal Investigation (CI) mission reads: "Criminal Investigation serves the American public by investigating potential criminal violations of the Internal Revenue Code (IRC) and related financial crimes in a manner that fosters confidence in the tax system and compliance with the law." This mission is based on IRS's statutory authority and responsibility to enforce these laws and supports the overall IRS mission of "...applying the tax law with integrity and fairness to all."

Evidence: CI's mission is included in its annual Strategy and Program Plan. IRS's mission is included in its 2005-2009 strategic plan (see www.irs.gov). CI's powers and jurisdiction are contained in U.S. Code Titles 26, 18, and 31.

Does the program address a specific and existing problem, interest, or need?

Explanation: The tax gap -- the difference for a given year between taxes legally owed and taxes actually paid - for 2001 (latest available figure) is estimated to have fallen between $312 and $353 billion. CI is one of the major IRS programs intended to minimize this revenue loss.

Evidence: See IRS's 3/28/2005 press release (www.irs.gov) "New IRS Study Provides Preliminary Tax Gap Estimate" (IR-2005-38). All IRS programs serve to promote voluntary compliance with the tax law. However, CI, Examination and Collection are the major IRS programs for directly interacting with potentially non-compliant taxpayers. Both Examination (2005) and Collection (2002) have been separately PARTed.

Is the program designed so that it is not redundant or duplicative of any other Federal, state, local or private effort?

Explanation: CI has sole jurisdiction over its primary mission, criminal violations of the Internal Revenue Code. This mission area uses the bulk of its staff. CI shares jurisdiction with other agencies on some financial crimes (e.g., FBI, DEA, Immigration and Customs Enforcement). However, it has a history of cooperating effectively in these areas including its participation in the Joint Terrorism Task Forces, the Organized Crime Drug Enforcement Task Forces and Corporate Fraud Task Forces. Often, when other substantive violations cannot be proven in these joint efforts, CI is able to bring tax and/or money laundering indictments to ensure task force success.

Evidence: CI's powers and jurisdiction are contained in U.S. Code Titles 26, 18, and 31. IRS has sole jurisdiction over Title 26 cases which make up the bulk of its workload.

Is the program design free of major flaws that would limit the program's effectiveness or efficiency?

Explanation: CI underwent a major reorganization in 2000 partially in response to a major external report (the Webster Commission Report) that determined that CI had moved away from its core tax compliance mission. Since that time CI has reemphasized tax cases, particularly abusive tax schemes, refund crimes, non-filers, and employment tax fraud. Further, in 2004 Professor Jeffrey Dubin completed a study that used a state-level cross-section time series dataset for the years 1988-2001 to model the relationship between tax payments and levels of IRS enforcement. His data supports CI's approach by suggesting that higher levels of criminal sentences lead to higher tax compliance.

Evidence: The Webster Commission report can be found at www.irs.gov/compliance/enforcement/article/0,,id=130626,00.html. CI's indictments based on legal income tax fraud cases have risen 17 percent from 560 in 2001 to 653 in 2004 despite lower agent strength. Narcotics indictments have dropped over the same period. A summary of Professor Jeffrey Dubin's study can be found at www.irs.gov/pub/irs-soi/04dubin.pdf.

Is the program design effectively targeted so that resources will address the program's purpose directly and will reach intended beneficiaries?

Explanation: CI has consistently high case acceptance rates from the Justice Department and has achieved high conviction rates. Further, more than half of its cases receive some publicity. This data suggest it targets its investigations well. In addition, CI works closely with Justice and other IRS operating divisions to ensure it targets efforts at high risk noncompliance.

Evidence: In 2004 Justice accepted over 90 percent of cases recommended for prosecution from CI and achieved a 92 percent conviction rate. CI estimates that 70 percent of its completed prosecutions received some publicity in 2004.

Does the program have a limited number of specific long-term performance measures that focus on outcomes and meaningfully reflect the purpose of the program?

Explanation: IRS' overall goal for all of its programs is to increase voluntary compliance as measured by periodic compliance studies. Its goal is to increase the individual voluntary compliance rate from the 83.5 percent level measured for 2001 to 85 percent by 2009. However, due to the time, expense and taxpayer burden involved in collecting data on compliance, IRS only periodically measures overall compliance. IRS annually measures the "Percent of Taxpayers stating that it is OK to cheat on their taxes" through surveys as a proxy measure for compliance. Criminal tax enforcement by CI is a key strategy in this effort and CI shares in these goals. IRS's CI specific long term goal is to "increase the number of convictions by 25% by the end of FY 2009 over the base year of FY 2002 while maintaining or exceeding a conviction rate of 92%." These metrics are used as a proxy for CI's impact on compliance. CI's impact on compliance cannot be directly measured, but academic research suggests that higher levels of convictions/criminal sentences lead to higher compliance.

Evidence: See IRS' FY 2007 Performance Budget for IRS' current strategic goals. CI commissioned an independent study by Professor Jeffrey Dubin which suggests that higher levels of convictions/criminal sentences lead to higher compliance. A summary of Professor Jeffrey Dubin's study can be found at www.irs.gov/pub/irs-soi/04dubin.pdf.

Does the program have ambitious targets and timeframes for its long-term measures?

Explanation: As noted in question 2.1, IRS's goal is to increase the individual voluntary compliance rate to 85 percent by 2009. It also has a proxy goal to decrease the "Percent of Taxpayers stating that it is OK to cheat on their taxes" from the 2004 level of 12 percent to 9 percent by 2009. Over the past decade, CI's workload has steadily increased in complexity. Tax evaders are becoming more sophisticated, making increased use of computers and electronic transfer mechanisms to move funds anywhere in the nation or to offshore tax havens. This makes IRS' CI specific goals for both the increase in the numbers of convictions (+25 percent) and the high conviction rate (92 percent) ambitious.

Evidence: See IRS' FY 2007 Performance Budget for IRS' current strategic goals. After dropping to just 1,926 in 2002, CI has raised convictions to 2,151 in 2005 mostly through increased productivity. Its long term goal will further increase this to 2,400 by 2009. CI has exceeded its 92% conviction rate goal has only twice during the past 15 years. CI's highest conviction rate was 92.9% in achieved in 1997. Its lowest was 87.9% in 1995. In recent years it has hovered between 91 and 92 percent.

Does the program have a limited number of specific annual performance measures that can demonstrate progress toward achieving the program's long-term goals?

Explanation: Starting in 2005, CI introduced three new annual performance measures that collectively provide a balanced picture of CI's performance. These include number of convictions, conviction rate and conviction efficiency. Number of convictions is used as a proxy for CI's impact on compliance. CI's impact on compliance cannot be directly measured, but academic research suggests that higher levels of convictions/criminal sentences lead to higher compliance. Conviction rate demonstrates the quality of CI's work since only properly investigated cases will lead to a conviction, and conviction efficiency (cost per conviction) measures efficiency. CI also shares in IRS's goal to increase the percentage of taxpayers stating that no cheating is acceptable.

Evidence: CI introduced these three measures in FY 2005 (see the measures tab).

Does the program have baselines and ambitious targets for its annual measures?

Explanation: CI's goals to increase the number of convictions each year while maintaining a high conviction rate and maintaining or reducing costs per conviction are ambitious. Further IRS plans to increase the percentage of taxpayers stating that no cheating is acceptable.

Evidence: See the measures tab for detailed performance metric data. After dropping to just 1,926 in 2002, CI has raised convictions to 2,151 in 2005 mostly through increased productivity. Its FY 2005 goal was 2,048, compared to a FY 2004 actual of 2,008. CI has exceeded its 92% conviction rate goal only twice during the past 15 years. CI's highest conviction rate was 92.9% achieved in 1997. Its lowest was 87.9% in 1995. In recent years it has hovered between 91 and 92 percent. Conviction efficiency (cost per conviction) is expected to remain below the FY 2004 level of $362,849 for the next several years.

Do all partners (including grantees, sub-grantees, contractors, cost-sharing partners, and other government partners) commit to and work toward the annual and/or long-term goals of the program?

Explanation: CI cooperates with numerous organizations in pursuing its mission. Its high acceptance rates for cases recommended for prosecution and for convictions demonstrate its successful partnership with Justice's prosecutors. It also shares goals with other agencies when it participates in joint task forces and shares goals as appropriate with other IRS divisions. This question has been given a lower weight than other questions in this section because CI's primary activity, tax related criminal investigations, is carried out by CI personnel.

Evidence: As noted in question 1.3, CI participates on several important joint law enforcement task forces. These task forces track their own performance. For example, the Organized Crime Drug Enforcement Task Forces uses "investigations tied to consolidated priority targets" as one of its metrics. Each participant contributes to these common metrics. See question 1.5 for case acceptance and conviction rates. When appropriate, CI and other IRS operating divisions share goals. For example, IRS is emphasizing referrals of suspected cases of tax fraud from its civil enforcement functions (examination). In the first quarter FY 2005 fraud referrals from other IRS divisions accepted (93) increased 52.5% over the same period in FY 2004.

Are independent evaluations of sufficient scope and quality conducted on a regular basis or as needed to support program improvements and evaluate effectiveness and relevance to the problem, interest, or need?

Explanation: CI's primary role is to deter noncompliance with the tax laws. Deterrence is difficult to measure. However, IRS recently completed a major tax gap study that measures overall IRS success in deterring noncompliance. CI is a major part of IRS' overall tax compliance effort. CI also commissioned an independent study by Professor Jeffrey Dubin to model the relationship between tax payments and levels of IRS enforcement. In 1999, the Webster Commission completed a comprehensive review of CI's operations and effectiveness. Finally, the Inspector General for Tax Administration and the General Accountability Office complete numerous studies on IRS effectiveness and management.

Evidence: See IRS's 3/28/2005 press releases for tax gap estimates (www.irs.gov). This tax gap data is based on audits of randomly selected taxpayers to project compliance for the entire population. A summary of Professor Jeffrey Dubin's study can be found at www.irs.gov/pub/irs-soi/04dubin.pdf. While his work was done under contract to IRS, it follows an accepted academic methodology and is undergoing peer review. The Webster Commission report can be found at www.irs.gov/compliance/enforcement/article/0,,id=130626,00.html. This was an independent commission chartered by former IRS Commissioner Rossotti. See also Inspector General for Tax Administration reports 2005-10-054, 2003-10-201, 2003-10-041, and 2005-10-081, and GAO- 05-707T and GAO-05-207.

Are Budget requests explicitly tied to accomplishment of the annual and long-term performance goals, and are the resource needs presented in a complete and transparent manner in the program's budget?

Explanation: IRS has presented an integrated performance plan and budget for a number of years. It has worked to tie resource changes to expected performance changes. In FY 2006, it proposed to restructure its budget to improve this linkage by showing the full cost of each program activity. Prior to 2006 overhead activities were funded separately. Unfortunately this proposed new structure was rejected by the appropriations committees in FY 2006 and IRS reports it cannot track obligations for these fully costed programs quickly enough to use them in financial execution. However, IRS will continue to monitor and report the costs of these programs. IRS 2007 Budget is presented in the traditional structure, which does not show the full cost of programs. However its performance budget also includes tables detailing the full cost of each program. IRS' efforts to link budget and performance is further improved by tying resources and annual performance measures to IRS's new long term performance goals.

Evidence: See IRS' FY 2007 Congressional Submission for its integrated performance plan and Budget. While the budget structure does not report the full cost of programs, CI's full resources are reported in supplemental tables with clear performance information and a discussion of the impact of resource levels on its annual performance.

Has the program taken meaningful steps to correct its strategic planning deficiencies?

Explanation: IRS has a strategic plan and processes in place to ensure that it focuses resources on the highest priority problems and explores process changes to improve performance. However, it needs to continue to improve its ability to measure the impact of its programs on compliance.

Evidence: CI works to continually identify and correct strategic planning deficiencies. It uses long term planning, annual budget and performance planning and regular performance reviews throughout the year. Reviews take place at all levels of the organization and take into account feedback from external parties such as the Tax IG. For example, these processes were used to identify problems with fraud referrals from other divisions and to take steps to address this problem (see question 2.5).

Does the agency regularly collect timely and credible performance information, including information from key program partners, and use it to manage the program and improve performance?

Explanation: CI collects data from internal and external partners and uses it to manage. This includes both data on CI's performance measures and a variety of diagnostic metrics to help managers understand performance trends. CI leadership is constrained by the IRS Reform & Restructuring Act of 1998 (Section 1204) which prohibits the use of "measures of enforcement results" to evaluate employees. However, within those constraints it collects and uses performance measures. CI has had some problems in the past with the timeliness and accuracy of management information. However, it is taking actions to address these issues.

Evidence: CI provides case completion and fraud referral data monthly to the Commissioner. This data shows CI and other IRS function progress in enhancing the Fraud Referral Program. CI also reports performance results, including convictions, in comprehensive quarterly program reports and meetings with a leadership team led by IRS' two Deputy Commissioners. These sessions are used to direct changes in programs and priorities and to report back to leadership on the results of changes from earlier sessions. CI also forwards monthly data to its prosecutor partners in the Department of Justice to help ensure coordination. CI's partners provide a variety of information to CI such as updates on case dispositions from Justice's tax division and leads from state tax agencies.

Are Federal managers and program partners (including grantees, sub-grantees, contractors, cost-sharing partners, and other government partners) held accountable for cost, schedule and performance results?

Explanation: CI's strategies and operational priorities are incorporated into formal annual managerial commitments. Each manager's performance is evaluated against these commitments annually. Further, IRS has implemented a pay banding program to increase the linkage between pay and performance for its managers. CI leadership is careful to ensure that these commitments do not violate the IRS Reform & Restructuring Act of 1998 (Section 1204) which prohibits the use of "measures of enforcement results" to evaluate employees.

Evidence: Recent CI manager commitments have included; emphasizing legal source investigations, increasing investigative efficiencies, enhancing the Fraud Referral Program, and improving the accuracy and timeliness of management information system data entries. Other IRS operating divisions have also included commitments related to supporting the Fraud Referral Program, a key CI priority, in performance agreements of their managers.

Are funds (Federal and partners') obligated in a timely manner and spent for the intended purpose?

Explanation: CI fully obligates it resources and follows its annual financial plans. In addition, CI has a Resource Investment Board (RIB) and is establishing a Technology Investment Review Board (IRB) to ensure that discretionary resources are invested strategically and within plan. This question has been given a lower weight than other questions in this section because CI is funded as part of an annual largely salaries and expenses account. This type of account does not typically have problems with obligating resources according to plans.

Evidence: CI reports that it obligated more than 99 percent of its allocated resources in 2002, 2003 and 2004 and fully achieved its hiring goals in 2004.

Does the program have procedures (e.g. competitive sourcing/cost comparisons, IT improvements, appropriate incentives) to measure and achieve efficiencies and cost effectiveness in program execution?

Explanation: CI's performance measures include a cost per conviction measure which has remained stable for several years despite annual pay increases. In addition, in response to the IRS Commissioner's directive to reduce cycle time, CI has taken steps to ensure that cases are closed as quickly as possible without compromising results.

Evidence: CI's primary efficiency measure is the conviction efficiency rate (cost per conviction), the CI budget divided by the number of convictions (see the accompanying performance measures table). Further, CI reports that average cycle time to complete cases for FY 2004 (397 days) was 9.8% less than the FY 2002 level (440 days).

Does the program collaborate and coordinate effectively with related programs?

Explanation: CI works with a variety of external and internal partners. It participates on several important joint task forces, coordinates closely Justice Prosecutors, operates joint efforts such as fraud referral with other IRS divisions and exchanges information with state tax agencies.

Evidence: Examples of successful collaboration with state and Federal partners include: - CI participation in a joint task force to which seized $6 billion assets of the former Iraqi regime, - A joint effort with the Austin Police which yielded $1.3 million in forfeited drug money and the seizure of 11 pounds of cocaine, and - Sharing information on CI's automated fraud detection system with New York state which helped New York build a similar system. See question 1.3 for information on joint task force participation. See question 1.5 for case acceptance and conviction rates which show coordination with Justice Prosecutors. See question 2.5 for an example of coordination with other IRS divisions.

Does the program use strong financial management practices?

Explanation: GAO reports that IRS "lacks reliable and timely cost information" needed to make management decisions (GAO-05-707T).

Evidence: See GAO-05-707T. GAO reports that IRS' new administrative accounting system may eventually give it reliable cost information. However, it will take several years to accumulate enough data. GAO (IRS's auditor) reports material weaknesses related to 1) financial reporting, 2) unpaid assessments, 3) federal tax revenue and refunds, and 4) information security (GAO-05-103). Per GAO; "IRS continues to exhibit a strong commitment to addressing its ongoing financial management problems and has made improvements in recent years that have resulted in the closing of many recommendations. At the same time, the continued existence of the serious financial management weaknesses that gave rise to the remaining open recommendations represents a serious obstacle that IRS needs to overcome to achieve effective financial management" (Highlights of GAO-05-393).

Has the program taken meaningful steps to address its management deficiencies?

Explanation: CI works to continually identify and correct management flaws. It uses long term planning, annual budget and performance planning and regular performance reviews throughout the year. Reviews take place at all levels of the organization and take into account feedback from external parties such as the Tax IG. The main management weakness identified in this section of the PART is ongoing problems with IRS' financial management, particularly the lack of "reliable and timely cost information" (see question 3.6). As GAO confirms (Highlights of GAO-05-393), IRS is working to address these problems by upgrading its financial systems. IRS's new Integrated Financial System (IFS) may eventually provide the sort of day to day which GAO reports that IRS currently lacks.

Evidence: Examples of CI's recent actions aimed at correcting management deficiencies include the reorganization of CI's IT sections, the addition of Assistant Special Agent in Charge (ASAC) positions, and consolidation of selected field offices. See question 3.6 for a summary of IRS' financial management issues.

Has the program demonstrated adequate progress in achieving its long-term performance goals?

Explanation: As noted in Section II, IRS's goal is to increase individual voluntary compliance to 85 percent by 2009. It also has a proxy goal to decrease the "Percent of Taxpayers stating that no cheating is acceptable on their taxes" to 9 percent by 2009. CI has made solid improvements in performance in 2004 and 2005 and demonstrated progress towards its CI specific long term goals. However, recently released data show that the tax gap remains large and survey data shows no improvement since 1999 in the percentage of taxpayers stating cheating is acceptable.

Evidence: IRS's CI specific long term goal is to "increase the number of convictions by 25% by the end of FY 2009 over the base year of FY 2002 while maintaining or exceeding a conviction rate of 92%." These metrics are used as a proxy for CI's impact on compliance. CI has increased both of these measures in 2004 and 2005 (see measures tab). After dropping to just 1,926 in 2002, CI has raised convictions to 2,151 in 2005 mostly through increased productivity. CI has exceeded its 92% conviction rate goal has only twice during the past 15 years. It fell slightly below this goal in 2005. CI's highest conviction rate was 92.9% in achieved in 1997. Its lowest was 87.9% in 1995. In recent years it has hovered between 91 and 92 percent. Finally, survey data showed 11 percent of taxpayers stated that cheat was acceptable in 1999. In 2004 this number was 12 percent (unchanged in statistical terms).

Does the program (including program partners) achieve its annual performance goals?

Explanation: CI exceeded two of its annual performance measures for FY 2005 (convictions and cost per conviction). However, it fell slightly below its 92 percent conviction goal achieving 91.2 percent.

Evidence: CI's performance measures were new in FY 2005. It fell slightly below its 92 percent conviction goal achieving 91.2 percent. It exceeded its targets for number of convictions (target = 2,048, actual = 2,151) and conviction efficiency (target = $332,194, actual $295,316). See measures tab. See also Treasury's annual Performance and Accountability Report for a discussion of FY 2005 performance.

Does the program demonstrate improved efficiencies or cost effectiveness in achieving program goals each year?

Explanation: CI's performance measures include a cost per conviction measure which has remained stable for several years despite annual pay increases. In addition, CI reports that average cycle time to complete cases for FY 2004 (397 days) was 9.8% less than the FY 2002 level (440 days).

Evidence: See the accompanying performance measures table for conviction efficiency rate, the CI budget divided by the number of convictions. Cycle time data above is captured by CI's management information system.

Does the performance of this program compare favorably to other programs, including government, private, etc., with similar purpose and goals?

Explanation: While there are numerous other law enforcement and tax enforcement programs, they are sufficiently different from CI that direct comparisons are either impossible or would not provide useful information.

Evidence: Other investigative programs also measure statistics like conviction numbers and rates. However, they investigate different types of crimes, so comparison of these numbers across programs is not meaningful. Similarly, other organizations enforce tax laws (states, foreign countries, other IRS programs). However, their circumstances are very different. In the case of other IRS programs, we have no means to compare their performance metrics or compliance impact to CI.

Do independent evaluations of sufficient scope and quality indicate that the program is effective and achieving results?

Explanation: Available evaluations present a mixed picture of CI's success. Professor Dubin's study suggests that higher levels of criminal sentences lead to higher tax compliance. Since CI has succeeded in raising convictions in recent years, this suggests some success. However, CI's conviction numbers remain low by historical standards. In addition, IRS's most recent tax gap estimates (for 2001) show that compliance has grown somewhat worse since it was last measured in 1988. CI has made significant changes to address criticisms from the 1999 Webster Report. Finally, the Inspector General for Tax Administration has issued several reports including criticisms of CI's operations (e.g., reports on CI efforts to refocus towards tax enforcement, the usefulness of its Lead Development Centers and coordination with civil tax enforcement functions), and GAO has included tax enforcement on its High Risk list.

Evidence: A summary of Professor Jeffrey Dubin's study can be found at www.irs.gov/pub/irs-soi/04dubin.pdf. See IRS's 3/28/2005 press releases for tax gap estimates (www.irs.gov). The Webster Commission report can be found at www.irs.gov/compliance/enforcement/article/0,,id=130626,00.html. See Inspector General for Tax Administration reports 2005-10-054, 2003-10-201, 2003-10-041, and 2005-10-081. See GAO- 05-707T and GAO-05-207.