CHAPTER FIVE: START-UP STRATEGIES

For most North Carolina communities, the main challenge of GIS is simply starting up. Geographic information systems are not cheap in time, personnel, or resources. Resource needs include ongoing budgeting for hardware, personnel, training, and technical assistance. A policy implication is that communities must be encouraged to have permanent budgets sufficient to allow the GIS specialist to use standard tools and thus benefit from collegial support within the crime mapping community. There is also a need to provide small to medium jurisdictions with technical information on optimal hardware specifications. Finally, there is a need to provide small to medium jurisdictions with technical information on data collection, reconciliation, and integration. Related general managerial precepts for GIS start-up are outlined in this chapter.

Strategy Checklist

Budget adequately and specifically

Smaller jurisdictions often have not yet faced up to having ongoing budgets which recognize information processing software as a recurrent expense. For instance, when asked why she was not using ArcView, a leading mapping package for crime analysis, but was instead using a package designed largely for home use, one crime analyst candidly admitted, "My agency doesn't want to pay for it." This denies tools to would-be crime mapping staff and is also a negative signal to staff regarding the importance of crime mapping. Likewise, funding must be adequate for ongoing personnel, hardware, data management, and training.

Given the special training and technical support needs of small and medium jurisdictions discussed in previous sections, and given the existence of nation-wide support and self-help networks of crime mapping specialists, which center on use of leading standard mapping tools such as ArcView, it is false economy for local communities to "save" a few hundred dollars by forcing specialists to use cheaper, functionally limited software which lack the same level of support within the crime mapping community. A common, desirable model is that pursued by the city of Wilson, NC, where GIS was funded from the city budget, with each department (police, fire, public works, etc.) also contributing to funding GIS operations.

Purchase adequate hardware.

Establishing a crime mapping system routinely involves a powerful system of networked computers (Nesbary, 2000: 20) and is not amenable to implementation on aging hand-me-down PC's. We asked, "One common problem is people trying to do something before they have the prerequisites to do it. Can you think of any examples of this happening in your agency or another crime analysis, or another crime mapping agency you know about?" The response from Austin, TX, was: ⁽¹⁾

"People get a grant and they say, oh, we want to start mapping out crime, and sometimes what happens is they underpower their hardware. A lot of people don't realize how processor-demanding GIS is, so a lot of what I've seen happen is they will underpower the hardware and not have it be big enough, powerful enough a tool to do the job."

Small jurisdictions, used to operating on machines suitable for word processing and low-level computing needs, may lack understanding of the necessity for more powerful hardware. They may be tempted to underpower GIS by settling for minimum requirements specified on the software box. Mr. Ed Orff of the Colorado Springs Police Department personalized some of the inevitable frustration arising from attempting to perform GIS operations with underpowered hardware:

"You'll have problems with the software losing data, and it's real frustrating to see people work 4 to 8 hours a day, and then at the end, I can't save it, or it locks up several times during the day because your machine is not fast enough to handle it."

Establish interagency cooperation

To avoid "reinventing the wheel" is one of the compelling reasons for establishing a system of interagency cooperation around GIS. Taking advantage of existing geographic databases is perhaps the most-recommended start-up strategy encountered during this study, as illustrated by the study's Las Vegas, NV, respondent:

"Get access to everyone else's geographic databases. If there aren't sufficient supplies of geographic databases in your area, that contain spatial data that you can use, persuade your local government agencies to create them. That's pretty much their obligation."

Mobilizing a municipality to invest in GIS may well require interagency cooperation over time, as the case of Winston-Salem suggested:

"What Winston-Salem did was go over to the Board of Aldermen, and get a CIP (capital improvement plan), so we could have some funding to really get an infrastructure in place, and I think that really did help us, rather than relying on individual departments' budgets. It just wouldn't have happened without that ... [The] Information Systems [Department] is where GIS was administered. We got together across multiple departments, with the Forsyth County assessor's office, looking at different primary departments in the city ... city/county planning, inspections, the IS, police ... we talked about what we wanted to do and ended up getting a $300,000 CIP (capital improvement plan), in order to invest in hardware. That was about '93-'94 when we started our efforts. The police department really had the software in play, had the equipment, but wasn't doing anything. [It was not until a crime analyst with GIS skills] moved over here in '98, and that point, that's where we could say we had a quarter of a person focused on crime mapping."

Smaller jurisdictions need to benefit from the experience of larger ones in avoiding the mistakes associated with underpowering their GIS systems. Software manufacturers' minimum hardware specifications are not a suitable guide for hardware acquisitions for purposes of establishing GIS functionality.

Anticipate database acquisition costs

Small to medium communities that already have base maps in some other function such as tax assessment provide an essential, must-have resource for crime mapping implementation. Conversely, communities lacking base mapping will encounter a substantial start-up cost in law enforcement efforts to implement crime mapping. From the Lenexa, KS, interview:

"I would say that the most difficult part of implementing crime mapping in smaller agencies is the availability of quality base map data ... quality base maps. Again, I was real lucky that this city had a GIS department that was already very well established, which made it very easy for me to hit the ground running ... [Elsewhere in my state] there are very few agencies that understand what base maps are, or that have access to good base map data."

Where base mapping is absent, start-up costs can be high. From the Austin, TX, interview:

"Then there is set-up, too. That's something smaller departments may have to think about, data cleaning and data integrity, probably are going to be more of a factor for a small to medium department than it is with us. Also, small and medium departments may have to go buy their street files from somewhere and that's another cost factor. They don't want to use the Census TIGER files, I never have used them but I've heard problems from other people that I've interviewed in the past."

That is, reliance on "free" Census street files may be the only feasible alternative for small jurisdictions, but in may involve other costs in terms of greater need for data cleaning later on. Likewise, as early efforts in Winston-Salem, NC, uncovered, over-reliance on GIS databases used by the municipal tax office are not always the answer either:

"When we first started trying to geocode [match police report incident data with map locations by address], and we did not have a centerline [street address database] going to the parcels, knowing that half of the addresses aren't really there in the database because for the ones that were not revenue sources, there was no location ... you're not very successful if all you're going to get is a 40 to 50 percent match rate."

Support data management needs

Failure to support basic data management functions leads to shortcomings in law enforcement effectiveness. From the Winston-Salem, NC, interview:

"We were annexing a piece of property in the city and I wanted to know what kind of crime was in that area, but [county agencies] couldn't tell me. A lot of agencies just turn in the uniform crime reports, or turn over reports to the SBI for them to tally; and they do no analysis themselves. For small to medium departments, I think data entry would be the biggest problem, because somebody's got to put that data in there. This means you're going to talk about personnel. After grant funding, when you get personnel, unfortunately the county and city commissioners, or whatever type of government they have, very rarely support those positions after the grant is implemented ... You need to look at long-range planning on how you're going to get that data and get a data entry program first, before you go into mapping."

Smaller jurisdictions need the benefit, if not of regionally centralized approaches to data management for crime mapping, then at least the benefit of profiting from the experience of large jurisdictions in the problems and pitfalls of establishing a high quality database, without which any level of sophistication in use of GIS software is meaningless.

Chapter summary

For small to medium jurisdictions, start-up strategies begin with needs assessment and proceed to financial planning for grants and long-term local funding, as in the case of the city of Winston-Salem providing a capital improvement plan (CIP) to build a sufficient GIS infrastructure. Start-up involves much more than just planning for adequate hardware. Planning for interagency collaboration around data management issues is central to what must be addressed from the outset.

Notes

1. Likewise, Ed Orff of Colorado Springs, CO , illustrates the same point:

"…have the proper hardware to go ahead and support it, as far as size, speed of the computers, because when we came back, we found out that our computers were not big enough to handle the software and the demands that it placed on memory…"

The Austin system at the time of interview used a dedicated quad processor server with a half gigabyte of memory and a second dual processor server used as a database server. Austin's nine analysts used 333-450 MHz PCs with 64 MB RAM, 17" monitors. Upgrades were planned.