An Index of the Moment

My colleagues and I at the National Center for State Courts and elsewhere have been thinking long and hard about a “justice index” here in the United States, as well as a similar index, the Global Court Performance Index (GCPI), applicable at an international scale. Though not unschooled in ways and means of performance measurement and management in justice systems, we’ve been daunted by the challenges that the construction of such indexes present.

What should the conceptual framework of the index be – the rule of law, justice systems institutions, legal frameworks, the experiences of citizens with the justice system, and so forth? Should the indicators that comprise the index be drawn from actual performance data available from justices systems, such as case clearance rates and median time in criminal defendants spend in custody before trial, or should the data be drawn from secondary sources, or should new measures be identified and defined? How much weighting should be given to the various indicators that comprise the indexes? How much vetting of the indexes is needed before development and deployment is initiated?


The Shoe-Thrower’s Index


Just do it! Aim, ready, shoot! That’s the bracing but refreshing lesson that I draw from The Economist’s use of an index of unrest in the Arab word that “aims to predict where the scent of jasmine may spread next” (The Economist, February 12, 2011; Economist.com/arabunrest). No doubt respectful of the challenges facing the construction of such an index, the Economist Intelligence Unit, the in-house research unit of The Economist that also publishes the Democracy Index, was not daunted. As a friend of mine said under similar circumstances, “I don’t know what’s good or bad at this stage, but I do know what nothing is!”


In its February 12 print edition, The Economist ran a table showing scores for 17 member countries of the Arab League using an index of unrest, or “Shoe-thrower's index.” Yemen came out on top with a score close to 90 on scale of 100, the most unstable. Libya, Egypt, Syria, and Iraq follow, each with scores close to 70. Qatar turned out the least unstable among the group with a score of a little more than 20.


Taking on the Challenges


The Economist intelligence Unit constructed the index by putting together a number of indicators that it felt were associated with instability – corruption, the age of the population, and so forth -- and ascribing different weights to them. (Where did these come from?  What august groups vetted them?) It removed a few of the members of the Arab League from consideration, including the Palestinian territories, Sudan and Somalia for lack of data, and Comoros and Djibouti, which do not have a great deal in common with the rest of the group. Some factors were discounted because they “are hard to put a number on.” The data on unemployment, for example, were too spotty to be comparable. The calculation of the index score ascribed a weighting of 35% for the share of the population that is under 25; 15% for the number of years the government has been in power; 15% for both corruption and lack of democracy as measured by existing indices; 10% for GDP per person; 5% for an index of censorship and 5% for the absolute number of people younger than 25.


Firing at Target Instead of Constantly Aiming


No one among us would dare oppose careful thought and analysis, but I do worry about an affliction that might be called the “aim, aim, aim syndrome” wherein one remains ever short of solution. Though one would expect an international news magazine to move more quickly than non-profit and public think-tanks and research institutions, one can’t help but be impressed with The Economist’s willingness not just to take aim at the challenges of an index of stability but actually to fire at the target with a solution, knowing full well that it might not have hit the mark perfectly, and that it might have to aim and fire again. In fact, in its online version, The Economist noted that Jordan comes out surprisingly low on the index, and suggested that the weighting “might need to be tweaked.” It ureged Interested readers to make suggestions and assured them that refinements will be made.

© Copyright CourtMetrics and the National Center for State Courts 2011. All rights reserved.

Aggregationists and Non-Aggregationists Unite!

Technology advances will dramatically change the way we do performance measurement and performance management (i.e., the use of performance information in  managements) over the next decade in what I call the "third wave" of court performance measurement. Writing in a special issue of Public Administration Review, the Urban Institute's Harry Hatry predicts that over the next decade managers and their staff will have an enormous amount of performance data at their fingertips.  They will be able to drill through various layers of highly aggregated data to disaggregated data, and slice and dice that data at will. Executives, managers and staff will have access to real-time cross-tabulations of outcomes for numerous variables.  For example, service user satisfaction outcome data might be calculated and displayed for various combinations of variables including identity of user (e.g., litigant, witness, juror), case type that brought users to court, gender, race/ethnicity, age, and location.

The ready availability of this enormous amount of performance information changes the way we construct  and use multiple performance measures. A case in point is the development of performance indices. Though not as robust in the United States, the development of governance, accountability and rule of law indices is a crowded field at the international level.

One way to achieve simplicity and yet monitor more than a few measures is to combine multiple measures into a single, simple index. Multiple measures in a "family" of metrics can be assigned weights according to their importance and combined in an index that is an aggregate statistic. Indices are attractive because they make complex information understandable and comparable across time and other places (e.g., courts, divisions, and locations).

The approach of aggregating individual measures to produce a high-level index – and the rating, ranking, shaming and praising that are brought into service by such an index – have sparked a debate pitting those who champion high-level indices against those that dislike them. On one side are the “aggregationists” who find indices attractive for the reasons noted above. On the other side are the “non-aggregationsists” who see risks that indices mask important differences, hide problems of data quality, decrease precision, amplify measurement error, and increase misinterpretation. They see no meaning in an index and find it useless for purposes of reform. (The characterization of this debate as aggregationsist versus non-aggregationsist comes from a July 2008 report to the World justice Project, Developing Indicators to Measure the Rule of Law: A Global Approach, by the Vera Institute who itsellf has weighed in with 60 rule of law indicators.)

Performance dashboards and business intelligence render this debate moot. There's no need to choose sides. Aggregationists and non-aggregationists can live in the same analytical space at the same time. They can move rapidly from a high-level index and drill down to highly disaggregated information, and back again via a different path at will. Performance dashboards are becoming the preferred way organizations – including an increasing number of justice sector institutions -- monitor, analyze, and manage their performance. They are the “face” of business intelligence designed to deliver the right information to the right people at the right time.

© Copyright CourtMetrics and the National Center for State Courts 2011. All rights reserved.