Doing What Works

The Obama Administration’s favorite think tank, the Center for American Progress (CAP), has launched a new project, Doing What Works.” Led by Reece Rushing and Jitinder Kohli, it has three objectives:

  • Eliminating or redesigning misguided spending programs and tax expenditures focused o priority areas such as health care, energy, and education;
  • Boosting government productivity by streamlining management and strengthening operations in the areas of human resources, information technology, and procurement; and
  • Building a foundation for smarter decision making by enhancing transparency, performance measurement, and evaluation.

At their kick-off event, they released a pair of reports:

Doing What Works: Building a Government That Delivers Greater Value and Results to the American People,” which lays out some specific steps the project will take:

  • Challenge the status quo.  Highlight misguided spending priorities and tax expenditures that should be eliminated or reformed, with particular attention to CAP priorities:  health care, energy, and education.
  • Measure what works.  Showcase examples of what a stronger measurement and evaluation system might look like to guide policy and management choices.
  • Experiment and innovate.  Test different approaches to improving results and apply lessons to other programs and agencies.
  • Coordinate and consolidate.  Identify federal programs that perform similar functions or serve the same people, and find ways that they could be better coordinated or consolidated.
  • Enlist the public.  Thousands of extra eyes can be employed to spot problems, offer solutions, and bring fresh perspectives. 
  • Be ready to execute.  Identify best practices to improve basic government systems, such as  information technology investments, contracting procedures, and the hiring process.

The report notes that such efforts are necessary because “a serious discussion of fiscal choices will only be possible if there is greater confidence that scare public resources will be wisely spent.”

The second report, “Golden Goals for Government Performance” offers five case studies on how to establish goals to achieve results.  The five case studies – the state of Victoria (Australia), Virginia, Scotland, the state of Washington, and the United Kingdom – each created an institutional focus on achieving broad societal outcomes.  Their experiences offer a possible path for the federal governmental performance model:

  • Define clear “outcome” goals at the highest level feasible.
  • Focus on the handful of goals that can only be achieved through collaboration across agencies.
  • Make sure public servants and the public are clear about what the goals are, and who in government is responsible for delivering them.
  • Build strategies to achieve these goals.
  • Evaluation agency programs on the basis of the contributions they make toward these goals.
  • Make the data on progress toward these goals transparent to the public.
  • Make the effort a joint one between the executive and legislative branches.

The case studies are both inspiring and revealing.  The report is worth reading.

Using Performance Measures

The federal government’s chief performance officer, Jeff Zients, declared at a recent Senate hearing: “The test of a performance management system is whether it is used.”  He thought federal agencies were failing the test.

So what should agencies do?  Well, last week, the Association of Government Accountants (AGA) released a research report last week that offers practical and concrete advice based on a recent study,“State and Local Governments’ Use of Performance Measures to Improve Service Delivery.”  According to AGA’s Hal Steinberg who led this effort, “This project was intended to determine how performance measures are used by governments to improve service delivery and also to describe their efforts in such a way that other governments can adopt similar practices.”

The study identified the key elements of performance management that contributed to service improvements.  It validated these elements via a series of case studies of four local governments (Baltimore, MD; Fishers, IN; New York City, NY; Westminster, CO) and Washington state government, as well as a broader survey of 175 other localities.

Key Findings.  Not surprisingly, the study concluded “. . . success appears to depend on the commitment and involvement of a chief executive who sees the process as a tool for improved performance, and not just a compliance activity.”

Other key success factors included:

  • the use of consistent measures from period to period to sustain attention on the measures, at the same time recognizing that measures can and should be modified when necessary to reflect changing requirements or expectations of stakeholders
  • regular and frequent analysis of the performance results data in comparison to prior periods, targets or benchmarks.
  • regular reviews of the analysis and results by the chief executive and/or his designee with the responsible agency heads, and
  • agreements with the agency heads on steps to be taken when the data reveal the need and opportunities for improvement.

Essential Elements. The research identified via the case studies several essential elements that must be present in a governmental system before it is likely to use performance measures to improve service delivery.   These were validated via the broader survey:

  • chief executive commitment and involvement in the overall process
  • relevant measures of at least outputs and eventually outcomes
  • periodic review and revision of the performance measures
  • frequent, regular collection of performance data
  • comparison of performance data to prior periods
  • regular review and analysis of performance results to ascertain the reasons for less-than-desirable performance and identify the opportunities for improvement
  • chief executive and other senior management participation in the reviews
  • agreement between chief executive and department managers on improvement plans
  • follow up on progress (or lack thereof) of improvement plans

Supporting Elements. In addition to the essential elements, the study also identifies several elements that survey respondents thought were important practices:

  • explicit targets for the performance measures
  • frequent comparisons of performance data to targets
  • comparison of performance data to the corresponding data from similar jurisdictions
  • intragovernmental comparisons (in states and larger jurisdictions)
  • support staff involvement in the reviews of performance results
  • budget reviews and deliberations considering the performance targets and results
  • budget resources allocated to programs rather than solely to object classes
  • programs for obtaining and considering constituents’ views
  • some process for assuring the data’s reliability
  • regular external dissemination of performance results

AGA concludes:  “This report demonstrates that the use of performance measures has already been adopted by many governments. It also provides guidance with which other governments can start to follow suit. The most significant missing piece is the desire and discipline to start and maintain the process-and that is ultimately up to each government body.”

Using Czars to Govern

Czar Nicholas IIThe media, and some members of Congress, continue to focus on President Obama’s use of “czars.”  An article today by the Wall Street Journal’s Neil King examines how this dust-up highlights the ongoing challenge of how government is increasingly facing problems that reach across traditional agency and program boundaries.  These problems include food safety, climate change, and the Recovery Act.  There is no accepted institutional mechanism to manage these problems and President Obama is using a pragmatic approach – appointing someone to be in charge on his behalf.

The media is focused on this particular management tool – the use of “czars” – and not the management challenge of how do you act on problems that cut across boundaries.  Professor Don Kettl predicted this growing conundrum in his book, “The Next Government of the United States,” where he notes that: “The importance of boundary spanning . . .  suggests a new approach for government – an approach that democratizes the process by spreading participation, privatizes government by relying more on nongovernmental partners, governmentalizes the private sector by drawing its organizations more into strong public roles, and ultimately challenges the framework of American democratic institutions.”

This is a real governance challenge.  But it is one that past presidents and other countries have faced as well.  As a result, there are several different models, other than “czars,” for addressing cross-cutting challenges:

Reorganize.  The traditional model is to reorganize government agencies and programs around a common mission or outcome.  This approach is cumbersome and does not readily reflect a continually changing world.  Examples of this approach in recent years:  the Department of Homeland Security and the Office of the Director of National Intelligence.

National strategies.  About a decade ago, the White House began to use “national strategies,” mainly in national and homeland security policy areas.  Under this approach, the President issues a “strategy” that provides an overarching strategic approach to addressing a particular cross-cutting issue.  Examples include:  the National Homeland Security Strategy, the Pandemic Flu Strategy, the National CyberSecurity Strategy. Sometimes someone was in charge (a “czar?”) and sometimes it was just a framework.

Performance-Stat.  Also about a decade ago, state and local governments adopted an approach initially piloted in the New York City Police Department called “CompStat.”   CompStat is a management approach that regularly brings together top leaders and managers to work regularly together to use fact-based information to address operational issues reaching across agency boundaries.  This approach, under various names, has been adopted in places such as: Baltimore (Citi-Stat), Maryland State Government (State-Stat), and Washington State (Government Management Accountability and Performance).  It could be adapted to the federal level, as well.

Delivery Unit.  Developed in 2001 by British Prime Minister Tony Blair, this approach is a variation of Performance-State, but focuses on strategic issues.  Blair identified 20 key outcomes that he wanted to achieve in his term in office, set measurable targets around each outcome, and charged his staff with ensuring strategies were in place and progress was tracked.  This approach has been recently adopted in the State of Maryland to supplement its StateStat efforts.

President Obama could use a mix of these approaches, including the use of czars.  He has already designated a “chief performance officer” who would be the logical focal point for coordinating these various cross-cutting initiatives.  However, this does not address one of the concerns raised by Congress:  how does Congress ensure accountability?

A Congressional Performance Resolution?  Congressional jurisdiction is organized around the traditional agency and program paradigm.  As Kettl notes, this paradigm is increasingly ineffective in addressing non-routine governmental functions.   So how can Congress be an effective player?  One recommendation offered by the Government Accountability Office is for Congress to adopt a “performance resolution,” similar to the existing budget resolution.  A performance resolution would be organized around major national outcomes.  It might parallel presidential “national strategies.” But it would, like the national strategies, define who is responsible for action on different elements of a broader agenda.  It also means that Congress would not have to reorganize its committee structure in order to become an effective player in cross-governmental challenges.