As noted in last month's feature article, "PART FY06: Positive Trends, Sizeable Exceptions," the Program Assessment Rating Tool (PART), having just passed its third anniversary, has experienced forward growth, but still has a way to go. This month's article addresses real variations in PART performance among agencies, and offers some takeaways for government executives.

How Agencies Compare
Some agencies view PART as just another regulatory obligation, while others embrace the process as a strategic tool for allocating precious financial resources. A review of PART activity at the agency level shows some major themes and great variety in both execution and performance.

1. All Agencies Are Not Equal
   
   Five agencies (HHS, DoD, SSA, OPM and Agriculture) make up nearly 75 percent of the FY06 funding captured by PART. The remaining funding is spread across 43 agencies. This is roughly equivalent to the proportions these agencies make up in the overall budget request (66 percent).
   
   
   
   Two organizations show a notable disproportion between overall funding and PART funding:

   • Treasury: 16 percent of 2006 budget authority; 1 percent of PART funding.
   • OPM: 2 percent of 2006 budget authority; 7 percent of PART funding.
   
   
   
2. All Programs Are Not Equal
   
   According to OMB's guidance, a PART program should have the following characteristics:

   • Be a set of activities recognized as a program by the public, OMB, and/or Congress;
   • Have a discrete funding level associated with it;
   • Correspond to the level at which budget decisions are made.

   
   How individual agencies have interpreted the guidance and determined the PART unit of analysis (what programs are chosen and how programs are defined) varies considerably. There are 13 agencies or departments with a significant number of PART programs (more than 20). Within this group, the average program size varies 7,656 percent, from over $10B at DoD to just over $100M at Interior.
   
   
   
   Some of the variation can be accounted for in the relative difference in agency size. For example, DoD's FY06 budget is 40 times the size of Interior's. However, even accounting for program size relative to agency budget, there is considerable variation.

   

   This variation has significant implications as agencies decide how to allocate analytical resources to PART; assuming the time to prepare and review a PART assessment is relatively equal, it will take HHS twice the time with PART as DHS and State, and three times as much as Commerce. Larger programs' definitions allow agencies to spend more time on individual PART assessments, notionally resulting in better scores and information.

   However, an artificially large program can cause confusion and the need for reconciliation between how agencies approach the budget and PART. For example, DoD's Shipbuilding Program is a $10B program (DoD's average size) in FY06. Internally DoD, manages shipbuilding through at least 40 individual programs (ACAT I-IV). Since neither DoD nor Congress typically develop budget, execute financials, or focus management at the shipbuilding level, the information contained in the Shipbuilding PART requires rework and has limited utility.



3. PART and BPI Ratings Are Not Equal

   Demonstrating PART proficiency is one of the key elements to getting to green in the PMA's budget and performance integration (BPI) component. As OMB points out "BPI is much broader than the PART and the PART is simply a tool used to achieve the much larger goals of the initiative." Still, since PART is the most measurable and visible portion of the BPI, one would expect a high degree of correlation between agency performance and PART and its BPI rating. However, analysis demonstrates an inconsistent correlation between them. There are 24 agencies that have both FY06 PART assessments and a BPI rating. Of these, 18 agencies also had FY05 PART assessments, which allow a comparison to the BPI progress rating in the PMA scorecard.
   
   
   
   Examination of the data points out several unusual cases. Most notable are differences between PART and BPI in the Corps of Engineers, with a red BPI rating and an average PART rating of 62 percent, and in the Department of Labor, which has a green BPI rating, but a worse average PART score. Additionally, four agencies (Education, HHS, DoL, and Treasury) received green ratings for progress in BPI while their PART averages declined in FY06.

Takeaways for Government Executives

While PART is still developing and has its challenges, it raises the stakes for strategic management in government and only increases in importance moving forward. Government executives should consider five key points related to PART:

• Agencies are doing reasonably well with PART assessments, although the real focus needs to be on how the PART unit of analysis, "program," is defined as the definition has real impact on an agency's ability to benefit from PART.
• Executives should keep an eye on the initiatives associated with PART that are on the horizon and prioritize PART efforts accordingly. It is likely that PART will become an indispensable tool to communicate the value of programs to both OMB and Congress, especially when comparing like programs.
• The link between PART performance and program funding decisions is becoming more concrete, but it is not yet definitive.
• There are no data to show that PART selection equates to funding cuts. When comparing programs selected for PART assessment and overall agency funding, PART program funding is just as likely to increase as decrease. However, more granular analysis is necessary.
• While PART impacts BPI, achieving BPI success, and green ratings from OPM, requires a broader focus.