|
Bang for the Buck: City of Tucson Real Estate Division Leaps Ahead with ArcView By David Koss and Cheryl A. Karrer Abstract The
City of Tucson Real Estate Division automated sixty 36” x 36” hand-drawn Real
Property Inventory maps. It would have
taken 1000 man-hours to manually reproduce the maps. COTRE created the new maps
in days using ArcView, the existing Pima County parcel base and street network,
and an HP 755 plotter. The cost to the
department for the ArcView 3.0 license was less than 5% of the Division’s
equipment budget, which is less than 0.1% of the Division’s gross budget. The money saved from this and other projects
undertaken is estimated at $200,000 - $250,000: a return on investment of over
250%. Introduction
The City of Tucson, Arizona is located
in the Sonoran desert region of the southwestern United States, approximately
100 miles south of the capital city of Phoenix, and 60 miles north of the
International border with Mexico. Tucson is a rapidly growing metropolis with a
population of over 500,000, and many booming adjacent suburbs.
The primary
responsibility of the GIS section of the Department of Transportation Real
Estate Division is to track an inventory of over 4,000 City-owned real
properties. The initial GIS goal within
the division was to reproduce a set of hand-drawn Real Property Inventory maps
showing the city-owned parcels. ArcView
proved to be a quick, easy, and inexpensive way to accomplish this, without an
insurmountable learning curve.
Figure 1: Dave Koss From Simple Beginnings… Prior to 1997, the
Real Estate Division used a Microsoft (MS) Access database to organize
information on the 4,000 city-owned parcels.
At that time, GIS within the division consisted of approximately 60
pages of 20-year-old base maps, with City-owned parcels drawn in by hand. Newly platted subdivisions and other updates
were made using the “pencil and eraser” or “cut and paste” methods. The maps
were never comprehensively updated; consequently, the pages were “dog-eared”,
torn, and unreadable. (See Figure 2) To
overcome this, a protocol for easily updating and reproducing this map set was
needed. ArcView provided the solution.
(Figure 2: The old map set) The Real Estate
Division’s foray into ArcView began in 1997 with a visit to Pima County’s
Technical Services division for a short demonstration of the software. The City’s GIS efforts were very small scale
and dispersed at the time. The County
was miles ahead, with several up and running mapping programs, including the
Pima County Land Information System, a cd-rom containing a run-time version of
ArcView 2.1 (Data Publisher) and more than forty shapefiles. Most importantly, the County had created a
digital parcel base for the entire county, as well as an associated street
network. The division looked to the
support of Pima County’s GIS Manager at that time, Howard Ward, for guidance on
the semantics of linking the Real Estates Division’s Access database to the
parcel shapefile coverage. After review
of both databases, it was determined that the Taxcode field in the
Access database and the parcel field in the parcel shapefile contained
the same data, and could be easily joined in ArcView. The seed was planted, and the Real Estate Division purchased a
single license of ArcView. The first task after
Dave Koss had completed a local Introduction to ArcView course was to
plan the Real Property Inventory Map Reproduction project. The Tucson metropolitan area is made up of
more than 30 Townships, containing over 341,016 parcels. Because of the tremendous number of polygons
in the parcel shapefile, it was decided to divide the area into smaller, more
manageable pieces than had been used in the original Real Property Inventory
maps. Since the parcels are
relatively small and numerous in the center of the City, it was decided to
divide the map pages up into quarter townships. By working with only nine sections at a time, we could easily
plot the parcels onto a 36” x 36” page for inclusion into the replacement map
set. For the maps showing the outskirts
of the City where the parcel size is typically larger, the scale was increased
to show an entire township. The first
project file was set up to include all views, themes, borders and layouts that
would become standard throughout this effort.
(See Figure 3)
Figure 3. Project
Design Next, the County’s
parcel shapefile was joined with the City’s property database based on the
common field. The City property
database was imported into ArcView using an ODBC connection in the “SQL
Connect” menu. A field was created in
the City property database called “color”, and attributed accordingly for each
controlling department: “blue” for
parcels controlled by the Water Department, “red” for parcels controlled by the
Fire Department, “green” for Parks Department properties, and so on. The legend was then created using the
“unique value” option on the legend editor, and the standardized .avl was
saved. The final step entailed creating
the layout and the final “look” of each map page. (See Figure 4)
Figure 4. Final Map
Layout ArcView project files
were then created for each quarter township, thus creating a separate project
for each map “page”. The original
project file was copied for each quarter township or whole township, and
renamed to reflect the map location.
After the project files were copied and renamed, the original parcel
theme from the template project file was deleted and replaced with the correct
theme tile for that area. Finally,
layout legends were adjusted accordingly, and the maps were plotted using a
Hewlitt Packard 755 color plotter. The
project wrapped up as a resounding success, as the old map set was quickly
replaced. A cover page was added for
quick indexing. (See Figure 5).
Figure 5: The New and Improved Map Set Word traveled quickly
through the City departments, and several requests for copies of the map set
were made. There was no doubt within
the Real Estate division that the one copy of ArcView paid for itself many times
over in man hours saved related to this project alone, and that the resulting
benefit to the City was almost immeasurable. In early January
2000, the division contracted with a local consultant, TerraSystems Southwest,
Inc., to write an Avenue script that further automated the map production
procedure. (See Appendix “A”) The
script selects quarter township and whole township “tiles” out of the parcel
base shapefile and saves each tile as a new shapefile with the township, range
and section as the file name. The
script is run on a weekly basis using the most current version of the County’s
parcel base, and the updated tiles are automatically placed in the designated
directory. Each project then is
automatically updated when the .apr is opened, allowing for easy update of the
map set. The master Real Property
Inventory Parcel database is kept on a different server and updated by clerical
staff on a daily basis to reflect any new additions or deletions to our
inventory. Updated themes are posted on
the City’s intranet for use by Property Agents and other City of Tucson
departments. Staff continues to make
hand drawn entries on the hard copy map set, which is replaced only once every
six months due to cost of printing.
Single map pages with numerous changes are plotted more often, on an as
needed basis. The old pages are removed
and the new pages inserted in their place.
In the end however, the division has a fresh copy of the map set every
six months instead of every twenty years, and the labor effort is minimal. Making this Success Possible: The Good, the Bad, and the
Not so Bad… The Start of GIS in Pima County, AZ
The availability of
the parcel base in shapefile format from Pima County played a key role in the
City’s ability to easily recreate the Real Property Inventory map set. Thanks to strong inter-governmental
cooperation from a grass-roots level, data sharing and coordination have become
the norm, with the City accessing the City/County jointly supported “MARS”
server – providing the most up to date parcel data available. Recently, the City and the local Council of
Government (Pima Association of Governments) have contributed back to the
effort by sharing high resolution digital orthophotos for the Tucson
metropolitan area, which are also available on the MARS server in both ArcView
and ArcInfo compatible formats.
However, the foundation for GIS in the City of Tucson was laid by a Pima
County project a decade earlier. Pima County began its
adventure into GIS in 1984, with the passage of a $5 million dollar bond package
for “a computerized mapping system” to be installed in all public
libraries. Pima County Planning and
Development Services, now known as the Development Services Department, was
placed in charge of this venture. In
1986, Gene Trobia, who is currently the Arizona State Cartographer, was hired
as the Pima County GIS Manager. Mr.
Trobia named the effort the Integrated Mapping and Geographic Information
Network (IMAGIN) Project and kicked-off a multi-year needs analysis and
demonstration project. During the demonstration period 1988-1990, the IMAGIN Project
leased Arc/Info software and ran it on a variety of platforms, including a
Prime minicomputer, UNIX boxes from Hewlitt Packard and Sun, and a VMS
workstation from Digital Equipment Corporation. In 1990, a Sun system was installed and production of the GIS
database began. In 1990, it was
determined that the implementation of a County-wide GIS program would cost
between $12-$20 million dollars, which was not easily available due to the
economic and political conditions of that time. Five participating County departments agreed to support the
expanded program, only to withdraw because of the economic downturn. The program was left with no additional
money forthcoming, and the situation for GIS was bleak. County Management communicated that the GIS
program should become an “Enterprise-Fund” department, which meant raising all
operating funds from the provision of data and services. This seemed like a reasonable solution in
the beginning. The GIS group, now
called Engineering and Geographic Information Services, would hire staff,
convert all the data, and ultimately charge all other departments, utilities,
and other government agencies, such as the City, a portion of the cost to
create the digital data. The base
layers were sold to many local utilities, and several hundred thousand dollars
were raised, but by 1993, the paying customers had run out. At that time, the
City of Tucson and its various departments were not allocating any funds to
assist in the development and distribution of GIS data locally. Furthermore, the GIS community was agitated
that a bond-funded program would charge for data. Many local agencies and departments continued to do business in a
manual mode, and some started redundant data conversion efforts. To make matters worse, the Engineering and
Geographic Information Services Department had a capital and operating budget
of approximately $1.8 million dollars per year and were running a heavy
deficit, even with the large data sales that had been made. Furthermore, due to a downturn in support
among various players within the community, GIS development was slowed for a
few years. Fortunately, the story does not end here. In 1995, the
Engineering and Geographic Information Services Department was folded into the
County Department of Transportation as the Technical Services Division. This move was done to link what was
perceived by County management as a valuable service to a stable funding source
– transportation bond funds. With
pressures from data sales lessened, the new division implemented a “damage
control and GIS resuscitation program”.
In 1995, the Pima County Land Information System was born. The Pima County Land
Information System (http://www.dot.co.pima.az.us/gis/pclis),
fondly known as PCLIS, was the brainchild of Howard Ward, who became the GIS
Manager in 1994. He recognized that
selling the GIS data was not working, and that the local GIS community was
nearly nonexistent, or perhaps just stagnant.
The thought was to create a CD-ROM containing various GIS data sets and
a run-time version of ArcView 2.1 (Data Publisher), with a customized GUI
simplifying many general GIS functions for the inexperienced GIS user- all for
a bargain price of $65. In 1996, PCLIS
was in full production. The 300,000+
parcel base served as the base coverage, along with the associated street
network and forty plus other useful GIS layers such as rivers/washes and
topography. The GUI allowed for users
to search for parcel information based on the parcel ID number, the street
address, or the owner’s name. In
addition, the GUI provided for easy one-button map production and other
enhancements such as user-friendly spatial analysis wizards, which helped those
not familiar with GIS technology utilize the data. It was this application that first got the attention of the City
of Tucson Real Estate Division. PCLIS
made data readily available and helped them to see how GIS could be applied in
their department. Today, PCLIS
continues to be updated on a quarterly basis, and the dataset count has grown
to over 150 shapefiles. Pima County
followed up its initial CD effort by moving data to the Internet. In early 1997, current GIS Manager Jack
Lloyd acquired a beta version of AutoDesk’s MapGuide and quickly built a basic
data display and query interface.
Since then, Pima County's MapGuide site (http://www.dot.co.pima.az.us/gis/maps/mapguide) has evolved
into an internationally recognized GIS Internet site. Hundreds of jurisdictional staff have come to rely on the County
MapGuide site in their everyday work tasks, including the Real Estate Division. The Division publishes the most current
version of the Real Property Inventory dataset to the MapGuide site on an
as-needed basis. As County, City, and
other jurisdictional data continues to be added to the site, the user base
continues to spread into the private sector and the general public. In addition,
digital orthophotography products are rapidly being incorporated into projects
throughout the County. Its applications
include landuse/land cover identification, building improvement assessment,
floodplain mapping, roadway design work, spatial enhancement of existing GIS
layers, generation of new GIS layers, facilities mapping, terrain analysis,
cartography, and a host of others that spring up every day. In short, it has been
a long, winding road that has brought the local jurisdictions, including Pima
County and the City of Tucson, to where they are today with information
technology. The creation of the
county-wide parcel base was crucial to the Real Estate Division’s Real Property
Inventory Map Reproduction project, and cost the County an estimated $400,000.
This effort was accomplished through intergovernmental cooperation on
all levels, and provided a platform from which regional GIS could grow and
flourish. Identifying and Quantifying the Benefits from the Purchase
of ArcView
The goal of
determining how to replace and easily update the hard copy Real Property
Inventory maps was obviously the largest single benefit from the purchase of
that single ArcView license by the Real Estate Division. Another benefit derived from the use of
ArcView was that the staff was able to quickly identify and capture parcels
that were somehow omitted from the property inventory database. This added many thousands of dollars in
assets to the City’s real property inventory, thereby increasing the City’s net
worth and strengthening their bond rating. The success of the
Real Property Inventory Map Reproduction project also led to requests by other
City departments and divisions for map products and related services. One such
undertaking was the Substandard Rental Housing Assessment Project. The Real Estate division was called on to
assist with a special project to assess and identify all substandard rental
properties in the city. A team was
gathered together from several different City departments to assess the rental
properties, but the problem remained as to how to identify the rental
units. ArcView and the County’s parcel
base were once again used to identify rental units based on a field included in
the parcel dataset that classifies property by type. The resulting theme was broken down by City Wards and maps were
quickly and easily created to allow assessment teams to plan their routes. The
teams went out twice a week and the entire assessment of 90,000 rental units
was completed in less than six months.
To contract this project out to the private sector would have cost the
City many thousands of dollars. Again,
the success of this project was based on the purchase of one license of
ArcView, and on access to data that had already been created by Pima County and
the City. From the various GIS
projects completed within the Real Estate division, it became clear that GIS
was becoming an integral part of the Division’s operations. More interesting however, was how the use of
GIS technology began to spread throughout other City departments. The timing may have been coincidental, but
perhaps grassroots movements begin with the success of a single project, much
like the Real Estate division’s example with the Real Property Inventory Map
Reproduction project. This project
demonstrated how one successful application of the GIS technology can promote
an entire department to be more efficient and productive, while also reducing
costs. This can then permeate
throughout other departments, ultimately changing the local governmental
landscape. Quantifying the benefits of
acquiring GIS technology can be difficult however, and the continued growth of
GIS within an organization is often contingent upon communicating with the
decision makers the benefits of this technology. Many department heads need to be convinced that GIS can be a
cost-effective solution to many problems faced by local governments. Conducting a cost-benefit analysis, as has
been done for this project, can help to establish a wider audience in which to
tout the possibilities that GIS can offer.
Showing a positive return on investment related to the acquisition of
GIS software can convince those who are skeptical about acquiring this type of
technology that it can be fiscally rewarding. ***add space here**** Cost-Benefit Analysis, ROI and Their Role
in GIS Strategic Planning: The Proverbial Bang for the Buck — Cost-benefit analyses
entail defining the costs and benefits of a particular venture and assigning a
monetary value, and are often conducted during the GIS strategic planning
process. A key attribute of a strong
strategic plan is a clear definition of the desired outcome from acquiring GIS
technology. A cost-benefit analysis
provides a means “to organize the
sometimes vague desires of a firm or agency to ‘use GIS’ in a solid, strategic
plan for implementing and running the GIS” (Korte, 2000). However, a cost-benefit analysis does not
necessarily have to reflect an entire GIS.
It may focus on “a specific output, application or a firm’s use of GIS
technology as a whole” (Wilcox, 2000).
To many organizations, whether considering the acquisition of GIS technology or
looking at the outcome of a specific project from an existing GIS, one of the
most important results from a cost-benefit analysis is the return on investment
(ROI) figure. The ROI figure generally
represents the comparison, presented as a percentage or ratio, of the benefits
versus the cost associated with a particular product, application or
expenditure. Defining Cost
and Benefits
There are many approaches one can take
when conducting a cost-benefit analysis, whether for an entire GIS system to be
acquired in the future, or for an existing system with the goal of determining
the cost-benefit of a particular product.
Recently, research conducted by Webster and Lombard (1999) focused on
applying traditional methods used by the American Society for Training and
Development (ASTD) to the evaluation of
training results in GIS, particularly the ROI.
Their ROI model was based on that of Dr. Jack Phillips (1997), which is an
augmented version of the standard Kirkpatrick method that was developed in
1959, and which is still considered the most widely used method of evaluating
training programs today (Webster and Lombard, 1999). This template was roughly adapted to determine the ROI for the
implementation of GIS technology within the COT RE Division, particularly the
concept of hard and soft data (Phillips, 1997). Phillips notes that data utilized in ROI calculations can be
divided into two categories. Hard data
consists of the traditional measures of an organization’s performance such as
items produced, project timelines, employee overtime, and overhead costs,
including equipment and software. Soft
data are more difficult to quantify such as work habits and climate, employee
attitudes, new skills, and initiative (Phillips, 1997). It is often difficult
to apply ROI to GIS technology. In the
past, costs and benefits often were considered over several years, since the
initial expenditures for hardware/software, staff training and application
development tended to far outweigh the definable benefits when converted to a
monetary value (Wilcox, 2000). As was
seen by Pima County’s experience, initial capital outlay for hardware,
software, and data conversion can be substantial, and the benefits difficult to
quantify. However, times and technology
have changed considerably since the early eighties, when GIS was run from a powerful
yet expensive Unix box. Today, GIS has
moved to the desktop, thus reducing costs considerably. Perhaps just as important, GIS has become
more user-friendly, leaving the command line behind for window-based GUI’s and
wizards. Also, data availability has
increased substantially, and thanks to intergovernmental cooperation such as
that between the City of Tucson and Pima County, data creation, maintenance,
and sharing has become a joint effort.
In response to all of these movements, benefits generated from newly
acquired GIS capability are now more quickly realized, sometimes seemingly
instantly, as was with the City of Tucson Real Estate Division. Governmental
departments or divisions often acquire GIS technology in a manner that does not
allow for a formal strategic planning process to occur. In many cases, other departments or
governmental entities in the community have already gone through the initial
stages of technology acquisition, which may or may not have included a
strategic planning process, and are actively applying GIS in daily
operations. Other departments that are
currently seeking GIS capability jump on the bandwagon without conducting a
departmental specific cost-benefit analysis, in an effort to leap forward into
the GIS technical arena. This can lead
to a quagmire of problems, because the initial venture is not well
defined. To remedy this situation,
there needs to be a clear understanding of the project goal, data needs and
availability, and hardware/software demands.
This does not always require a lengthy or forma strategic planning
process. Such information can be
gleaned from an informal, less structured manner. Sometimes this
informal approach is taken after the fact.
After the acquisition of a new GIS, it is often useful to conduct a
cost-benefit analysis “to confirm or correct the course of action” (Korte,
2000). That is, to readjust if needed,
before acquiring any additional components, and before committing any
additional resources. The City of
Tucson Real Estate Division and its automation of the Real Property Inventory
Maps is an excellent example of how conducting a cost-benefit analysis and
determining the associated return of investment figure after the successful
completion of an initial project can help define the role of GIS within the
department, and secure future funding for expansion. For the City of
Tucson Real Estate Division, an ROI analysis was performed after the completion
of the initial project using the Phillips’ technique of data classification and
categorization formats propounded by Darlene L. Wilcox in her paper entitled GIS
Implementation (Wilcox, 2000). Ms.
Wilcox broke the analysis into GIS relevant categories for the cost and benefit
data requirements. From this example,
categories where then modified to reflect the situation at the City (see Tables
1 & 2). ***bump the following portion to next page – keep tables
togetner on same page – each table should have its own page***
(Adapted from Wilcox,
2000) The use of
less-than-precise estimates, assumptions, and external data sources may concern
some GIS managers, making them hesitate to conduct soft data conversion to
monetary values, or even attempt to apply this process of ROI evaluation. In order to minimize these concerns and
raise credibility, Phillips (1997) suggests: à
Take a conservative
approach when making estimates and assumptions à
Use the most credible
and reliable sources for the estimates à
Explain the approaches
and assumptions used in the conversions à
When the results
appear overstated, consider adjusting the numbers to achieve more realistic
values à
Use hard data
whenever possible Assumptions
Used for ROI Calculation
It was
estimated that it would have taken over 1000 man-hours to manually reproduce
the Real Property Inventory maps, resulting in a set of maps that were not
easily updateable, and which would be out of date before they were even
completed. However, reproducing the
maps manually was not a likely scenario given that electronic base coverages in
GIS format did exist for the area.
Therefore, this factor was not incorporated into the cost-benefit
analysis.
Hard Data
Figures
The hard data
figures used in the cost-benefit analysis were determined from City of Tucson
Real Estate Division gross budget numbers and known GIS-related
expenditures. See table below for
specific data references. ****ditto****
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
