Previous: Contents

Up: JudgeIt A Program for

Next: History

Introduction

JudgeIt implements several statistical procedures intended to help evaluate electoral systems and redistricting plans. It carries out a number of other functions as well, such as generation of descriptive statistics and graphics, that are useful for related purposes. JudgeIt also provides means for social scientists to evaluate a large number of characteristics of electoral systems, their fit to the data, and the adequacy of any predictions made. The methods used by JudgeIt to evaluate electoral systems and redistricting plans are based on concepts developed in our scholarly work, listed in the references at the end of this manual.

The program can evaluate electoral systems in three general situations:

1. When an election already has taken place,
2. When an election has not been held yet but a new redistricting plan (or plans) has been proposed or implemented, and
3. When you wish to assess what an election would have been like if held under certain specified counterfactual conditions (such as if no minority districts had been drawn, or term limitations had prevented incumbents from running for reelection).

For example, JudgeIt will estimate partisan bias, the extent to which an electoral system favors one party, and electoral responsiveness, the extent to which party control of legislative seats responds to changing party success at the polls. These two concepts have been used in the academic literature for some years now, and partisan bias may be operational version of some of the concepts explained in Davis v. Bandemer 106 U.S. 2797 (1986), the Supreme Court case that declared political gerrymandering justiciable but only gave outlines of what an acceptable measure of it would be. JudgeIt will also estimate and graph seats-votes curves, make specific vote and seat predictions for individual districts, and calculate numerous other relevant statistics.

A particularly important feature of this program is that, for bias, responsiveness, seats-votes curves, and virtually every other estimate, JudgeIt provides quantitative estimates of uncertainty (i.e., standard errors). Although all good social scientists (and even the Supreme Court, see Castaneda v. Partida, 430 U.S. 482, 1977, and Hazelwood School District v. United States, 433 U.S. 299, 1977) recognize that presenting some assessment of uncertainty is crucial for any empirical analysis, too few studies of electoral systems and redistricting plans do so. We hope that by making the calcuation of standard errors easier, JudgeIt will help change this situation.

As described in our scholarly articles, we developed the methods used here over many years, evaluating them with data from hundreds of congressional and state legislative elections in tens of thousands of electoral districts throughout the United States and elsewhere. We have developed some extremely sophisticated versions of our models, with refinements for numerous features common in legislative elections data. However, our most recent set of modifications led to a surprising but useful result: many of the difficult parts of the statistical model fell out algebraically, leaving a much simpler mathematical form. One particular advantage of this current approach is that the methods implemented in JudgeIt are based primarily on regression analysis, the most common statistical technique used in the social sciences. The technique therefore should be fairly straightforward for any social scientist to understand and use.

For details on the methods used here, we encourage users to read the article listed as Gelman and King (1994), ``A Unified Method of Evaluating Electoral Systems and Redistricting Plans,'' and possibly some of the earlier articles cited in the references (see page ). Although we cite only our own work here, since JudgeIt implements the methods we have developed, one should not overlook the numerous other scholarly articles and books written on these and similar topics by hundreds of researchers in half a dozen disciplines over the past century. Citations to some of this extensive literature can be found in our articles.