|
Ten Things My Mother Didn't Teach Me About Redistricting
By Steve Miller, Chief
Wisconsin Legislative Reference Bureau
Delivered at the NCSL National Redistricting Seminar: Plotting the 00s Maps
Denver, Colorado
March 5, 1999
This paper comes from the point of view of a non-partisan redistricting staff that works for all members of the Legislature. My remarks do not necessarily represent the views of either the Mississippi or Wisconsin Legislatures.
Ten years ago, I had never been involved in redistricting--but I was 15 years younger then. I became staff director for Mississippi's redistricting committee in 1988. I suddenly needed to learn many new things. I started attending the NCSL Redistricting Task Force meetings, where I learned a great deal. The Task Force, through its meetings, network and publications, offers the best source of learning about redistricting. Then I learned even more by doing redistricting. Today, I will share some of the things I learned on the job. I will focus on the ten things my mother didn't teach me about redistricting.
1. Know your Mission.
This will help you stay focused on the real business of redistricting. In Mississippi, the mission is: "Draw a legislative redistricting plan according to constitutional standards by December 3, 2001." Read your statute and your state constitution--everyone else will. For example, the Miss. statute authorizes the committee to commandeer staff from other state agencies. It also limits travel reimbursement for committee members to 30 days.
Does your mission include service to the public? Probably not, but some members will want you to act as a public relations office for the Legislature's redistricting effort. You should consider putting a site on the World Wide Web to disseminate basic Census data, precinct maps, redistricting plans, and information about the Legislature's redistricting process.
As part of knowing your mission, you should understand your history. If you're new to redistricting, it will help you to find out what issues emerged in the 80s and 90s redistricting cycle. You may see these again in slightly altered form.
2. Make Hard Decisions Early.
Our first decision was to write our software locally. We did not start planning our system until Christmas 1988. That was too late. Mississippi faced statewide elections in 1991. Four states, including Mississippi, had elections in 1991, and had to draw their plans during the same year that they received the census data. Mississippi received its census data on January 31, 1991, and we started drawing new districts the next day. We had to design and build the airplane as it taxied down the runway. Due to careful planning, we achieved liftoff.
Time pressures caused a lot of stress. Here are some of the things we did early, that made our job easier:
- created a large scale map book showing every member's residence (by census block);
- set up a public-use terminal (which didn't get used very much);
- printed demographic reports of the 1982 redistricting plan and the current voting precincts;
- set a cut-off date for adding changes in local precincts to our data base (October 1, 1990);
- adopted eight criteria to govern drawing of all plans to be considered by the Committee;
- calculated the actual cost of maps for sales to the public, as required by law;
- bought an "E" size color plotter and photocopier (we also used small plotters extensively);
- started a dialog with local officials (we sent out informational memos);
- worked out a process with local officials for getting digital precinct maps;
- adopted rules for acceptance of third-party plans;
- hired outside legal counsel to assist the Committee; and
- made provisions for members to keep plans on floppy disks.
One useful device is a time line, or a PERT chart.
In 1994, we looked back on our experience and planned our purchases of equipment and software in stages up through the year 2001. We slightly modified this plan each year. We added compactness testing and contiguity testing to our software design. Due to recent Supreme Court decisions, you should consider whether your software will include a "compactness test." Computers can calculate indices of compactness by several recognized methods.
Early on, you should select the features your software will offer, and you will have to pick what data to include in your files. Will you have past electoral behavior of each precinct? For which elections? It seems like just a data problem to the staff, but election history data is very subjective stuff with complex political implications.
A big part of the job is getting data from local governments. You may need current precinct maps and election returns. If you get election data, you must also have precinct maps in effect on the date of the election.
3. Provide Security for Files.
You should secure both your paper and digital files. As a general rule, you should copy all data on a computer disk to backup tapes daily. A disk crash won't hurt too badly if you have yesterday's plan on the tape backup. Redistricting plans generated by legislators are the very essence of the process, and you must treat them as extremely valuable documents.
You should also plan a uniform method for labeling, filing, and indexing plans and maps that you generate and that you receive from others. You can create a form that has the minimum information that should appear on the plans and maps.
4. Redundancy is Good! [Hire lots of good staff]
When you're hot, you're hot--and everybody loves you. Legislators will love you. They'll come to see you every day, all day long. They'll call you at home. You will get to know them well. You will need more staff. We discovered that redistricting resembles "tag-team" professional wrestling. The Legislature's team had 174 members, and the staff's team had about ten.
You need people to gather and edit data, maintain the software and network, draw districts, and keep up with all the collateral files. You should plan a way to get extra terminal operators on short notice. You must train the staff to run redistricting software. Plan for enough staff to cover for illnesses.
5. Redundancy is Good! [Get plenty of good equipment]
Make sure you have enough computer equipment. Never underestimate the need for more terminals, better processor speed, more disk space, faster networks, or color plotters. Equipment will break down.
When legislators take the time to sit at a computer terminal to draw new districts, speed really counts. Spend time early in the project to make sure that equipment you buy will function adequately. How many active terminals will you need at a time? How long will work sessions last? We usually had two terminals going at once, usually for about 12 hours per day. But at times we had four terminals running, and sometimes we worked 16 hours a day for days at a time. You need fast computers and lots of terminal operators to support this effort. Don't forget to buy comfortable chairs.
One day in the office, someone said, "There are a lot of ways to skin a cat," meaning there are an infinite number of ways to draw a district. Legislators will want to see many different plans. We kept our terminals busy most of the time.
6. Separate the Warring Factions.
I don't want to give redistricting a bad name, but redistricting is sort of like war, and war is hell. You should know that doors and windows can cause problems. Separate staff from members--give staff some privacy. You should talk to Capitol Police about security issues.
7. Don't Play on the Railroad Tracks.
My mother probably did tell me to stay off the tracks. The lesson is to take extra care in the vicinity of great forces. No matter how many computers and staff you have, you will use them. In times of crisis, the demand will exceed the supply. Someone must act as gatekeeper to determine who gets to use the machines and the staff. Redistricting can be one of the toughest issues that a legislature has to deal with because it affects every member personally. I heard it said over and over again, "Redistricting brings out the worst in people." At some point, if the politics get too tough, the staff may tend to jump in between the contenders. However, you shouldn't feel the urge to step in between two oncoming trains.
Don't make promises you can't keep. In your eagerness to please and demonstrate competency, don't forget the high stakes involved, and that you lack control over the multitude of players.
8. Learn the New Math.
If you really try to draw minority districts--maximizing them--your minority districts will fall well below the ideal population. If you try to minimize the number of those districts, they will become overpopulated. Therefore, the average (mean) population deviation from the ideal district size for all minority-controlled districts can indicate the effort to create minority districts.
Consider the following equation: A = B / C
A = theoretical maximum percentage of minority-controlled districts in the plan
B = minority percentage of the overall population
C = percentage of a minority needed in a district for it to control the district
So, if you have 100 districts, the minority makes up 40% of the population, and it needs 60% of a district to control it, you could draw 66 districts controlled by the minority. This equation assumes an overall deviation of zero percent and ignores compactness. In the real world, you can draw a plan with a 10% deviation overall (for a state legislature), which may make it easier to draw minority-controlled districts. You cannot achieve the maximum because of the dispersion of the minority population, but you can probably do more than you thought possible. This formula explains why plans drawn by human beings tend to gerrymanders rather than compactness.
9. Adopt a Grand Strategy.
Sometimes the staff draws the plan, and sometimes the members do it. But whoever is drawing the plan, it helps to look at the state overall before you start.
In a statewide plan, where you want to make only changes that are absolutely necessary, you can "rubber-sheet" most of the existing plan. In an area experiencing dramatic population gains or losses, you may have to "pop" a district from one part of the state to another. But "popping" a district is a decision that is easier to make on the front end than after the plan is nearly done.
If you are working with a particular criteria in mind, start in the most important place. If you want to create minority controlled districts, you will start with them. Generally, the place where you start will have the most compact districts, and the last area where you work will look the most contorted. You can judge this book by its cover, because oddly shaped districts reveal the stresses of using criteria which conflict. If a district looks like a lizard, it probably is one.
10. The Real Game is Politics.
Never forget that technology is only a tool, and will play a minor part in the real game of deciding the geographical content or shape of districts. Redistricting is foremost a political activity, which is supported and constrained by technology and law.
Even the U.S. Department of Justice has a political agenda. It gathers comments from citizens, keeps them secret, and uses these statements in its analysis.
Summary
To run a redistricting office, you will need to understand the nature of census data, the technology you will use to manipulate the data, the political environment that surrounds redistricting, and the legal restraints upon the process. Add that to the ten things that I learned on the job, and you will succeed in redistricting.
|
