C SO. 23 7 : (* Geographic Base (DIME) m — A Local Program Conference proceedings November 18 and 19, 1974 Columbus, Ohio STREET ADDRESS RANGE ZIP CODE COUNTY CODE AREA CODE TRACT NO. BLOCK NO. DIR| NAME | TYPE LOW HIGH S Liberty AV 152101 152199 47233 035 010 001600 318 S Liberty AV 152201 152299 47233 035 010 001600 319 s Liberty AV 152000 152098 47233 035 010 001600 320 s Liberty AV 152100 152198 47233 035 010 001600 223 s Liberty AV 152200 152298 47233 035 010 001600 218 Light «• , „ >01 499 47302 03? 010 002101 UJ Light \ V^ 99 47302 035 010 002101 112 Light //"A cViip 5 *^' \^ >°~ , 47302 035 010 002101 HI Light A^fV Light f^J < ](; C\ \s^~~7/ ^jr* ->"> s, 035 010 002101 101 Sff*-. 5 '^ y^i%^ /. r 010 002101 201 Light v^£r r v>' / )o', l ^ / 1 »»/ 002101 210 Light ^X '^Srf; »« t ■''a Vi^ "\)>*S*J" ) 7 1 £/ / \ 211 Light (1%-^A'' '/ 1?i ^r" «°'l «// •~JJ 3 / 1 220 Light 1 NiXZ? Ls^^* ' s ^% 7 */ ffc/j" / a 221 Light Y'V'V?^ atC^ i/st/ 1 "1 // / */ ^?~*~> ~~^5;* 01 222 Light \y^>^ re: ^"^\ ^f^sii' »s * / .5 / *7 7 */ >7 u" -^^lOI 223 Lila<\ /&X^ " ' p\* f 't- ^r~^ «,] ^^^1/ °/- r /> s - 1 .yVVooo 402 IAla^C^/^ 12 ^ — -t-^-A"«»« V»^«09 ^ji// 1 ~ l s i ?/r2000 402 N ~^ !Xj ^^S-r^' '/\ f4^-\— "Os^^^i^f. ^7^*/* ^~4~Jl ;/ 'J1500 304 M LiR ^J^Q»\/ ! < / o "" 2"""-^ / */ Zl ° */ / - P^T: C01500 305 H Lil^V^ A/ , • ' "it\r / $/ ?/ ~ P^^iJ r J01500 306 N 14& v^v^x -7 / 1 ^Si*4 J o; ^7^\' / ^~4j7 /**«L«v 4001500 213 N L ' C "^ '°\J<./T r-5ifc i . a " V <^T~- ■i^i /~ "^^1 \ 5/ « 001500 212 N r \y Yv-x (?"/ \"~ ^»^~-g !NT> '"' /\/ 7^~ 001500 211 ^r°7% Computerized Geographic Coding Series GE 60 No. 6 U.S. DEPARTMENT OF COMMERCE Social and Economic Statistics Administration BUREAU OF THE CENSUS *«*« s of * Digitized by the Internet Archive in 2012 with funding from LYRASIS Members and Sloan Foundation http://archive.org/details/geographicbasediOOunit 1 >- a o o a TO SUB-SYSTEM DATA BASES Contiguous Block Block Associated with the segments are one or more blocks represented by the block record linked to the segment record. Any set of contiguous blocks may be linked together through that record providing the solution to many of the undefined boundary aggre- gation problems. We have found that running blocks instead of running or matching streets is more efficient in an interactive environment even though this sub- system allows both techniques. Through the various linkages, address to node to X, Y coordinate conversions are handled quickly, as well as X, Y to node to address. A standard point in polygon algorithm is utilized for that problem, but we see the undefined areal aggregation problem as the primary power of such a structure. A set of linkages to the three other subsystems provide the ability to extract data of various categories (i.e., crimes, financial resources for a particular pro- gram, and social/economic factors) for analysis and input into the next level of the planning process. Figure 2 shows a conceptual view of the entire I MIS including a retrieval and analysis subsystem. This allows the noncomputer person to compose questions, yet undefined, to the computer and receive replies, free of computer jargon, the results of a variety of analysis routines, or in graphic format. This subsystem is simply another way to represent and use the basic GBF data and to harness the power of its geocoding principles. One of the biggest problems in our minds was the communication of various kinds of geographically oriented data to nontechnical people in a form that would have a high probability of affecting the decision- making and planning process. Research, practical experience, and consultation with behavioral scientists convinced us that, indeed, a picture was worth a thousand words and was certainly of more value than 14 pounds of computer print out. We have designed and installed a multimedia informa- tion display system. A schematic diagram is depicted in figure 3. A variety of media including 35mm slides, micro- film from a computer operated retrieval unit, and computer generated graphic information can be mixed together and displayed on a large-screen (5 feet by 7 feet) color television projection system. This system can be used for both internal planning sessions or for communications to private citizens in public meetings. Further, because it is TV based, we have tied it into the cable TV network for either live or taped broadcasts. 83 FIGURE 2 INFORMATION REQUESTS Data Item Dictionary Retrieval and Analysis Locational Sub System Municipal Administration and Control Sub System Criminal Justice Sub System Community Development Sub System 84 FIGURE 3 Large Screen Video Projector 35mm Random Access Slide Projector CATV NETWORK Color TV Camera K> TO MICROFILM RETRIEVAL UNIT m High Resolution TV Monitor Graphics Computer Terminal TO COMPUTER w w In TO CATV NETWORK TO VIDEO TAPE RECORDER Out Video Mixer 85 At this time, we have too limited knowledge concerning all the things this kind of system can do. But again, an extreme amount of flexibility has been built into it and we are getting new ideas every day. We do know that it is an invaluable tool to close the communications gap between the public and private sectors in geographically oriented information. There are several systems which are or will be supported by these geographically oriented tools. One is a Community Activity Indicator Project. In coopera- tion with HUD and LBJ School of Public Affairs in Austin, Tex., the city of Tulsa has designed and implemented a set of community indicators that reside in the Community Development Data Base. These indicators are tools which make inferences about various physical, economic, and social conditions in our community. Problem areas can be recognized and resources can be allocated to remedy their condition. The indicators are not cure-alls, but as the word implies, are "an indication" of a condition in a partic- ular geographic area. In order to monitor the various broad-base areas of the community, the Indicator System has been divided into 1 1 categories. These categories represent a cross- sectional view of the total community. These cate- gories are: Sense of community Economic Education Environment Health and well-being Housing Land use and recreation Public safety Transportation City service delivery Demography These indicators, collectively, would generate a variety of information to infer conditions related to housing. Each of the 11 categories is described in the same manner. At the present time, the indicators are aggregated at a planning district level. These are homogeneous neighborhoods and do not necessarily coincide with standard census geography. Hence, the use of our Locational Subsystem comes into play by relating data elements by common geographic areas to be used to construct an indicator. As two simple examples, figure 4 shows how the indicator "physicians per 1,000 population" is calculated using outside medical data and standard census demographic population data with projected updating. Figure 5 shows an indicator which reflects intracity mobility. In both examples, data is aggregated at levels and areas different than those reported. The desired results were obtained through a geographic conversion scheme based on the GBF-based Location Subsystem. FIGURE 4. Number of Physicians per 1,000 Population DESCRIPTIVE VALUE Indicates the availability of doctors to the population. This indicator will show where the city stands in relation to other cities and States, but the indicator does not show whether all those who need the attention of a doctor are getting it. METHOD OF CALCULATION Total number of physicians x 1 ,000 Population Within each category, there are several indicators that comprise each particular category. For example, under the category of housing, some of these indicators are: Vacancy Single/multifamily building starts Number of demolitions Number of abandoned units Number of subsidized housing units DATA SOURCE, FREQUENCY OF AVAILABILITY, AND LEVELS OF DISAGGREGATION Oklahoma State Medical Association, Oklahoma Osteo- pathic Association, Yearly, Planning District/City. REFERENCE LBJ "Universe of Community Indicators," p. 48. 86 IGURE 5. Percent of Households in City Experiencing Intracity Mobility DESCRIPTIVE VALUE The indicator is proffered on the assumption that a community having a greater rate of mobility (intracity) will concomitantly have a lesser "sense of community." METHOD OF CALCULATION Total number "off" utility connections Total number of households DATA SOURCE, FREQUENCY OF AVAILABILITY, AND LEVELS OF DISAGGREGATION City of Tulsa, Utilities Services Department, Monthly, Planning District. rently it indicates the impact on liquid and solid waste, air pollution, school children, water supply, parking, and other ancillary factors. It has been designed to be modular so that other service delivery functions can be added as we develop the linkages. Again, this system, which has a large potential in the private sector, is geographically based. It must draw data from a variety of sources aggregated at different levels. The more complex the calculations, the need for a common geographic base grows. By maintaining a centralized, standard location system, we can alleviate many of the problems. Going full circle, once the indicator and projection systems have shown a need or problem in a particular area of the city, the planning process will develop programs to correct the situation. These, of course, are translated into budget terms in the next program budget. The locational subsystem is also linked to the Municipal Administration Data Base so we can com- pare the actual fiscal resources being applied to a particular area for a particular program for monitoring and evaluation. REFERENCE LBJ "Universe of Community Indicators," p. 117. Figure 6 shows a graphic representation for the indicator "communicable disease index". You can see that the new planning district boundaries do not coincide with census tracts or square miles. Another system which shows great promise for the decisionmaking process is called The Proposed De- velopment System. This is a projection system which is used to evaluate the impact of proposed developments on various municipal service delivery systems. Cur- I have tried to give some indication of how the basic GBF data and processes can be converted to an integrated data base orientation, and give you some ideas on possible applications. This is certainly not a static system and must be flexible enough to aid the ever-changing decision problems of local municipal management. With such new responsibilities as general and special revenue sharing bring, the local government must have better and more timely geographic data to develop commu- nity policies and evaluate their effectiveness. We believe that our data bases revolving around our centralized locational system give us a sound base to support out municipal management. FIGURE 6 Tulsa Metropolitan Area Planning Districts 87 ■■ DISTRICT BOUNDARIES COMMUNICABLE DISEASE INDEX - V.D., T.B., HEPATITIS - = ABOVE CITY AVERAGE = BELOW CITY AVERAGE '»'S» MITIOfOUHN *»t» PlANNINC CO**** IS MO 88 Question Period Mr. John Meyers— You mention the use of a cable as one of the possible subsystems to be used with the GBF. I am wondering what percentage of the popula- tion in Tulsa is wired for cable, and what kind of success you have had in obtaining citizen feedback via this cable on any particular problem put forth for mass response. Mr. Mitchell Tucker— The cable installation is still going on in Tulsa. The percentage of people signing up for it is very high, and as I said, that particular system is too new to get any kind of feedback at all yet. What we are finding is that the presentation methods using large screen TV in the room are better for com- munication. Dr. Robert Aangeenbrug— One of the difficulties we had in Wichita Falls 4 years ago leads us back into an issue which we may discuss later. In Wichita Falls it was pointed out that, theoretically, two-way cable TV's would be acquired. This resulted in a project discussion that really affected the staff. Do you know whether the superdesign in Tulsa also allows for two-way TV? We discovered that technically we could reach a point where a city could turn one's TV off because of an overdue water bill. Or, we could broadcast an individual message on the screen for delinquent utility bills. You can see how we moved, theoretically at least, into some very difficult policy questions. It brings back not too pleasant memories. Did any of that discussion take place in Tulsa? Mr. Mitchell Tucker— We do have two-way capa- bility, but in terms of turning them off if they didn't pay their bill, that has not been discussed yet. They keep raising their rates and that turns enough off. Mr. Alan Pisarski— I guess I found the whole discussion of your system very attractive. The last element, the proposed development system, I think is an excellent approach. An unnamed county in the Washington Metropolitan area got together with an unnamed consulting firm and tried to develop a cost benefit system for any kind of an initial development in the city, to find out what the total cost to the city would be, or the county, and what the total benefits in tax base would be. I will never forget the output of about a half million dollar-research program which came up with a sophi- sticated series of equations in a model structure, and the summary said that the design was conceptually correct, but there was not a data base anywhere in the world to feed it. At that stage they just dropped the whole subject. It seems from the way that you were describing it that you have taken a very straightforward and kind of clean approach to it. Is your approach based on existing rates of cost and output per unit input? Is that how you do this kind of work? In other words, the air pollution per unit of travel, or sewer cost per square foot? Mr. Mitchell Tucker— Each segment, each delivery system has a set of tables that are changeable, and that is really how it works— the capacity, for example, of our water system, which does change of course; the capacity of our sewage handling facilities, which can remain stable but does change as we add to the treatment facility capabilities, and so on. Mr. Alan Pisarski— How do you calculate the new demand that some development would create? Do you have simple rates to work with? Mr. Mitchell Tucker— There are a set of tables that we use in our calculating. We have two ways of doing it. One at the medium standard for this kind of development or the maximum of this kind of develop- ment. Again, it is not magic or anything like that. It is a pretty straightforward way of doing it, but it gives a pretty good idea of what is going to happen. Mr. James Thomas— What kind of utilization do you have for the community development aspect of this? What kind of demand do you have? How available is it? Who uses it? Mr. Mitchell Tucker— Are you talking about the proposed development or the community indicators? Mr. James Thomas— The community indicators. Mr. Mitchell Tucker— To give you some idea of utilization, the library has been deluged to the point that they want to put a computer terminal in the library for public access. We have had requests from everybody, from private companies to the Salvation Army. We are now battling, by the way, a policy problem of publication and distribution; it has been running our presses into the ground. We so far have been able to ward off many special computer runs. We have a policy that we will not be a service bureau. Any standard reports that can be published, extra copies for example, of course, are public record, unless it is confidential data. Mr. Alan Pisarski— Would you comment, in rough scale, on the cost picture over the years of your program. 89 Mr. Mitchell Tucker— The Mayor says it is too costly, and I say it is a bargain. These things are not, of course, a major development program, which is looking at all of the data bases and all of the conversion problems; massive amounts of data cost a lot of money, no question about it. We have utilized grants, but once the data bases are in place then these other systems come along rather inexpensively. Even the display system, itself, was rather inexpensive. This TV equipment is getting to the point that anybody can afford it now. My department, which does all the systems develop- ment for all the city, the Planning Commission, and the COG has a budget this year of $970,000 and that includes all of the computers, everything. The city of Tulsa had a population of about 335,000 in 1970. The SMSA was 502,000, that includes the three counties that are in the COG area. That $970,000 is a high. In fact, the systems I have been discussing are a very small part and the indicator systems did not consume much cost at all. I cannot tell you exact figures. The Proposed Development System has taken a long time and a lot of coordination, and really boils down to the time and effort of the people both in the public and private sector. They chipped in their own time, and the computer programing was really nothing. Mr. James Thomas— When you mention the popula- tion that you are servicing, is that the city of Tulsa, the 335,000, or is that the three-county area of 500,000? Mr. Mitchell Tucker— I am serving not only the 335,000 within the corporate limits, but also the three-county region, the 502,000. Mr. Alan Pisarski— I think we have had two really very interesting discussions. One of the things that we might play with during the discussion later on is the question of unit cost per capita for cities of various sizes and try to figure out if there is some magical number. I know in the early days of urban transporta- tion planning, we used to use a very convenient number of $1 per capita total cost for a large scale transportation study; that was in the very large cities. I do not know how well that worked in the smaller cities. I will be coming back to both you gentlemen during the discussion stage. I guess I can warn you now to comment on the nature of GBF/DIME building problems in your city and particularly with respect to whether the prime focus of the problem is institutional, financial, or technological, and perhaps look at it in that light. Market Area and Branch Location Analysis for a Large Commercial Bank Using GBF Systems ROY K.CAMPBELL Geoprocessing (data gathering and manipulation within a geographic framework) is rapidly gaining use and users. For many years geoprocessing applications were found principally in public and quasi-public organizations, but in recent years private business has become increasingly aware of applications of geopro- cessing technology. Commercial banks are among the largest, if not the largest, single group of private business users of geoprocessing. The structure of retail banking (individual and family banking needs) provides numerous opportunities for using geocoded data: market share and penetration analysis, product and service cross-sell pattern development, service area delineation, branch location analysis, and strategy planning for branch network evolution. Thus geopro- cessing of existing data sources can provide timely and spatially accurate input to actionable marketing and branching programs. The experiences of Manufacturers Hanover Trust Company (MHTCo.) are illustrative of problems which have occurred, or can be anticipated in, many business- oriented uses of geoprocessing as a marketing tool. MHTCo. is a New York City headquartered commercial bank with deposits of well over $26 billion, it ranks third by deposit size in New York City and fourth in the country. Ninety -five percent of its 1.8 million accounts are retail, and deposits in these retail accounts provide a key portion of the money necessary for the lending function by which a bank makes its income. These retail accounts are distributed among nearly 200 branch offices serving the New York metropolitan area and some upstate New York communities. The large volume of geographically-oriented data spread over an extensive area provides the prerequisites for geoprocessing, but also generates the first set of problems— availability, accessibility, and processing of internal data. Virtually all banks provide a variety of account types— regular and special checking, savings, loans. credit cards, etc. A banking unit (an individual, family, business, etc.) may have several types of accounts and, often, several accounts of the same type (for example, a special checking account, two savings accounts, a personal loan, and a bank credit card). At MHTCo., each account type constitutes a separate data base with between 50,000 and 1,000,000 account records; to further confound the matter, most account bases consist of two files— a name/address file and a ledger file, which contains balance and account activity information. Almost all marketing applications of these data bases require both address and balance informa- tion, thereby necessitating a merger of the two separate files into a single file containing name, address, and balance data. The uses to which marketing research wishes to put bank customer files suggests compression of the now merged account files into household, or more accurately, banking-unit records. This can be done for each banking unit by combining all accounts with the same last name using the same mailing address into a single record containing name, address, and account types with their respective balances. At MHTCo., compression reduces the original number of accounts by a third with each new "master record" representing a unique banking unit. This compression produces a high degree of analytical accuracy in cross-sell pattern research and cost efficiency in direct-mail efforts by minimizing duplicate mailings. Unfortunately, our production-oriented data processing center is not equipped to handle our file merging and compression needs, a fact which forces us to find an outside supplier, who must become familiar with the file structure and idiosyncrasies of each of the several account bases. This familiarization process is both expensive and time consuming, thereby placing a premium on finding a supplier with the capability and, more importantly, the expertise to process data beyond the simple file merging stages. Urban Data Processing, Inc., supplies the necessary file familiarity, software availability, and experience for our needs at 91 92 reasonable time and dollar costs. Our use of UDP illustrates a problem solving approach we've found most effective: purchase the solution to the problem whenever possible, rather than risk the time and budget overruns inherent in most developmental efforts. Another research application employing geographic base files is our ABLES (Automated Branch Location and Evaluation System). ABLES allows us to take an overview of an area, determine banking potential for each simulated market we postulate in the area, and rank them in order of most to least promising. Although the merged and compressed master bank- ing-unit file, called the Customer Data Base, serves many purposes in and of itself, its implicit research capabilities require additional, external data. Using an outside supplier to manipulate our large volumes of internal data allows our research and planning staff more time to become involved in the tasks of finding, evaluating, and using these external data sources. The various geographic base files for the New York metropolitan area present a case in point. Virtually all spatially related research and planning for branch banking activity require locational detail below the ZIP code level. The required detail is available through the use of address coding guides, which assign census and other area tie-in codes to a record based on the address in the record. While not a conceptually difficult task, care is necessary because inaccurate geographic coding can limit, or worse, distort conclu- sions drawn from research based on the data. The ACG published by the Bureau of the Census for New York City would be the most obvious source of census- related data; however, it has been found wholly inadequate for large-scale geocoding at the tract level because of an unacceptable number of errors and omissions. A reject rate of 20 percent would leave well over 200,000 records to be manually coded, which is not a trivial problem. Furthermore, while sampling might be appropriate for some applications, any research conclusions used as input to spatial manage- ment decisions would be questionable, because the extent of errors due to improper coding would be unknown. Rather than attempt to correct the Census Bureau address coding guide, we sought out an alternate source. The Reuben H. Donnelley Corporation main- tains a very accurate ACG for the New York area, which we were able to obtain as a part of a larger research project. After validating their guide and evaluating its limitations, we used it to geocode the Customer Data Base with excellent results. The cost of obtaining this file was minimal relative to the time and manpower saved over attempting to edit and improve the Census Bureau guide or to manually geocode numerous rejects. We also added several other descrip- tive data elements to our ACG, the most important for our purposes being the branch of MHTCo. into whose territory the particular street-address-range record falls. The ABLES computer programs combine four basic elements: census data on population and housing units, lists of commercial business, survey data on bank usage, and a file of all competing banking institutions. To a great degree, all of this data is integrated through the use of the bank's GBF— the Donnelly New York City ACG. Each household is assigned a banking potential, based on income. A Bank Usage Survey had been taken in an earlier year to gather detailed information on such banking habits as family income, number and types of accounts maintained, average balances, and how far people traveled to their bank. The survey data was broken down by area type (suburban, city residen- tial, poverty, etc.) and translated into a series of probabilities. For example, "the average" suburban family with an income of $12,000 has a .997 chance of using a bank for any service and a .946 probability of maintaining at least one savings account at a com- mercial bank. The average number of accounts is 1.66. Multiplying all the factors together with expected balance figures results in a "commercial bank savings potential" of $1,039.67 for a household. From the survey results, a table of potentials for income ranges was developed; these were translated into linear equations for a smoother progression. (That is, a man's balance would not suddenly jump several hundred dollars as his income changed from $14,999 to $15,000, or wherever the break occurred). The result- ing deposit potential of the blockgroup's "average household" was multiplied by the number of house- holds to determine the total residential potential of the blockgroup. Similiar logic was used to determine business poten- tial. A commercial list of business in the area, classified by size and type of each business, was obtained; previous studies had given an indication of what percentage of net worth was available to banks from different types of businesses. The individual business values were summed by type and multiplied by the appropriate percentage to determine total business potential. The final data element of the system was a computerized file of all competing banking institu- tions. Since an area might already be "overbanked" by 93 competitors, total potential would not give an accurate representation of an area's ability to support a new branch. Potential had to be adjusted for this compe- tition factor. From various reference sources, a record was established and verified for each institution; each bank was coded for status, type, geography, etc. All approved applications and existing banks were included in the file; different weights were assigned to savings institutions and commercial banks. The four data factors discussed above were aggre- gated into simulated market areas by means of a geographic coordinate system. Data was "collected" at a series of points; these points were "collected" into market grid totals according to the distances between them. The Medlist coordinate file was used for the population data. (This file had been created by the Census Bureau for the Office of Civil Defense. From street maps, a population center for each enumeration district was estimated, and a point plotted. This was translated into computer-readable form by using a digitizer; coordinates were attached to census data and released as a special file. Thus, the population data came "prepackaged"). The commercial business list had to be coded to the appropriate geographic unit by first computer-matching it against our address coding guide, and then hand-coding as necessary, to aggregate business potential to the same units as residential potential. Bank locations were precisely plotted on a map and then digitized; coordinates were combined with identifying information. The large area to be evaluated (a county, for example) was divided into grid squares; the size of these squares was determined by population densities and travel-to-bank distances that had been indicated in the surveys. Potential for a new branch at the center of each grid was measured by "collecting" all block group points within the prescribed distance and dividing by competition in the immediate area. (Diagram on page 00 illustrates this concept). The grid pattern was shifted several times to obtain a number of combi- nations of blockgroups. Finally, all of the individual grid totals were ranked to determine the parts of the county that showed the greatest potential and that should be more closely examined for new branch sites. An ABLES run for Queens County can be used to illustrate the points made above. Because it is a densely populated urban area, a grid size of one mile was agreed upon. (This meant that all "potential points" within one-half mile of the center of the grid, at which potential is measured, would be included in the grid totals). The competition area was defined as an "inner bank grid" of one mile on a side (same as the basic potential grid) within which competition would be given full weight, and an "outer bank grid" of one-quarter mile beyond. In this fringe area, competi- tion would be considered at half its usual value. Grid centers were established at one-quarter mile intervals to obtain as extensive a coverage as possible. Results were plotted on a map, (color-coded according to amount of potential) and showed clear patterns of which areas could probably support a new branch and deserved further investigation. ABLES is essentially a simple system. We have not tried to make it say "locate a branch on the North East corner of Front and Main streets", rather it says "look for a site in or around the sourtheastern part of Bayshore." Human judgement is heavily built in, permitting off-the-model evaluation of spatial anomalies involving rivers, cemetaries, parks, and the like. We've been able to update the Census population and income data through commercially available sources while keeping the bank file up-to-date our- selves. Since the Donnelly-derived ACG is only a little over a year old, we've not had to face updating it yet. I joined Manufacturers Hanover in 1968. One of my first projects was to find a method of relating the then upcoming 1970 Census data to the bank's marketing problems. I first heard of the address coding guide in 1969 and over the years, the concept of geographic base file, as realized in the ACG, has been a foundation about which our research people have erected a sophisticated approach to integrating customer records, external marketing data, and geographically packaged population information into analytical tools that have both saved the bank money and generated profits in their own right. We do not have a DIME file for New York City yet. The complexities of our ever changing metropolis are such that we may never have one. We'd certainly like to have one— the DIME approach is the "top of the line" in geographic base files and could be easily used in our pragmatic approach. The graphic capabilities alone stagger the imagination. We'll probably have to wait until a broader segment of private industry joins its weight (and dollars) with the academic and civil groups who are already knowledgable geographic prac- titioners, then the aggregate demand will cover the dollars which have not been supplied to date. 94 ILLUSTRATION: How Potential and Competition is Aggregated Map shows urban area with simulated market area of one square mile (inner grid). All block group coordinates within the inner grid are aggregated into the grid potential total; banks in the % mile outer grid, as well as the inner grid are considered as competition. 9 Hypothetical site for new branch bank X Location of competing bank • Block group centroid — - - Boundary of outer grid (for competition) ■■■■■■ Boundary of inner grid 95 Question Period Mr. Alan Pisarski— I think I would like to comment. There was very little in that presentation that I would say was unique to banks. If you are talking about service-area analysis, defining service areas around some activity, a supermarket, or a fire station for that matter, whether it's a public or private unit that provides a service, it seems to me that this kind of spatial distribution approach is still functional and still quite adequate. I think that the notion of a vast difference in the approach of the private and public sector is really not necessarily true. A great deal of site analysis that is done for public facilities where there is an optimization function that may be slightly dif- ferent, but not really terribly different, you are still trying to locate the best set of facilities in the best locations. Mr. John Meyers— Did this particular procedure hold true in your expansion to other areas in New York State, that is, outside the New York City metropolitan area? Mr. Roy Campbell-We have just taken a simpler approach. We have taken an area-analysis approach first. We are in your area now in Buffalo with a very small bank. The regulatory agencies would only allow banks like ours to take over fairly small banks, and most of them are located outside the urban areas. We are now in Buffalo, Rochester, Albany, down in Orange County. We expect to be in the Syracuse and Binghamton areas. The approach we take in these locations is to go in and in effect do an overview, kind of melting together the whole geographic economic approach to look at it. Then we work in small territories like that to the ZIP code level because data at that level is easier to work with. At that level we can certainly get the business data, especially down to a ZIP code territory, for anywhere in the United States using the ZIP code record. We had one of our fellows go out to the University of Michigan and get his doctor's degree, and he worked on a series of models checking bank usage, relating the 1970 Census with a big survey that was done by the Survey Research Center at the University of Michigan. He derived another way of approaching bank potential different than the ABLES file that we have found works very nicely in territories outside of New York City. Given the size of banks we have, we do not have to know all of the data in painstaking detail. In most cases they will only allow us to put up two or three branches a year until 1976 when it becomes statewide. We do not need to know the fine detail in areas like Buffalo. The business potential becomes considerably more important, so we run our business list very heavily in those areas. Conceptually it is pretty much the same approach; it is the idea of going from the top down. The particular details of application upstate are a little bit different. Again, the pragmatic has to come in, and we are looking at various cost constraints. It has worked very nicely, though; it has turned around our whole relationship with what we call the holding company clients in the division. That is our sales guide, to convince them that this kind of information is useful to them, and that it will give them a better concept of the ways they spend their dollars. It has worked very nicely for upstate. General Discussion Mr. Alan Pisarski— Some of the points that were raised yesterday deserve consideration. The institu- tional question came up again and again of how do you organize to make these things happen. It came up regarding two levels of coordinating mechanisms. With- in a city, within an urban region, how do you put together the set of players in some formative fashion, hopefully point them in the correct direction, and keep them pointed in that direction? But also, I think, the question arose of coordination between cities, among cities. How do cities become exposed to what is happening in other places? How do the developments in individual cities become exchanged? We talked about clearinghouse functions perhaps to be carried out by some of the existing professional associations, which I think probably popped into the minds of some people. I was absolutely astonished that we did not get into the discussion, at any time, of what do you do outside the area of existing addresses, or nonstandard areas. This morning we heard an additional series of papers on differing applications and experiences. One of the common threads that I noted in them was what I would call a very sensible sense of scale, modest efforts, not in their accomplishments, but in their pace. I did not see grandiose designs where the program design component takes all the money and 3 years. People started out working in a very pragmatic fashion. I have asked Mr. Morton Meyer from the Geography Division to join me on the podium, and Mr. Aangeenbrug, who will be wrapping up this conference at noon, to help lead the discussion. I will be calling on people representing cities here to perhaps get some sense of feel of their experience in their particular area. We would like to relate as much experience of different cities as we can so that we have some sense of levels of progress and the problems that are happening around the country. Mr. Charles Wenner— I want to express my gratitude to this group for coming to the city of Columbus. I have heard an awful lot of things, some I understood and some I did not. Of course, that is why we have people employed, that are here today from the city of Columbus, that can answer questions prior to decision- making problems. When I first started listening to this conference, I questioned why I should be here. It was not too long until it was very clear to me that if I have a problem or a question regarding computers I can pick up the phone and call Mr. Savoia, and he can come up with the answer. I think the benefit has been that it has made the people in the city of Columbus that work with this on a daily basis, cognizant of the fact that there are other people all over the country who have gone through what they are going through; and maybe, as indicated here yesterday and today, that Columbus is far advanced in this field. Maybe we can share some of the expertise that we have with other communities throughout the country. I would hope that as a result of this conference, the papers that have been produced might be made available to any of the next conferences that you might have because I believe a lot of information has been given out to the people. I am sure these same people will not be available at future conferences. If you could compile some of the history of conferences and where we are in Geographic Base (DIME) Files, and so forth, I think it would be a benefit for all concerned. I might also mention that there were a couple of words used yesterday and today that might be a barrier to what some people are trying to perform. One was "consolidation." Here in the city of Columbus, we have managed in the field of emergency medical services to share and cooperate with the surrounding communities. If you are in a remote point in the city of Columbus and the next community has emergency medical services, and a call goes out, they will dispatch the nearest emergency squad to serve you. This, of course, is a step towards regionalization, but it is not consolidation. Everbody wants to keep their positions, keep their own community, but it is working down that road; so the word "consolidation" is not too good. 97 98 The other word I heard was a "welfare bank," and I hope all those depositors that deposit there don't ever hear about that. Again, I do want to express my appreciation to the people who put this conference together and to all the people who participated. I think they all did a very excellent job, and we would be most happy to entertain any future conferences. Mr. Alan Pisarski— Perhaps I might ask some of the other people from Columbus— Mr. Henderson, Mr. Thomas— to talk a little bit about the issue that came up about how you consolidate and how you coordinate these activities. Perhaps more importantly, how do you get the word out to people that this tool exists, and let them know what you have got and what the capabili- ties are? Mr. Ronald Henderson-We have had a problem here in Columbus in getting the word out. We have been concentrating on maintaining the file. A few of us went to Indianapolis to talk to Mr. John Rowe. We were really impressed with how they concentrated on using their GBF/DIME, an uncorrected file. They conse- quently got their city-county government to fund the CUE program, because they realized the value. We have taken the other approach— maintaining the file. I think we have a better file to work with, but it is a little tough getting people turned on to using the file. We are concentrating on that right now. That is one of the reasons why we were really excited about this confer- ence—to bring some of our local public officials and private firms to this conference to show them what the GBF file really is and what can be done with it. We have plans for conducting seminars internally in our own agencies and also for the State of Ohio, to help bring all of Ohio's GBF/DIME's up to date, to share applications with everyone in Ohio and, of course, everyone nationally. We have had calls from as far away as Dade County in Miami from criminal justice people interested in our on-line address match- ing program. We have taken advantage of application programs from Mr. Rowe and his people. One of the things we do -not believe in is reinventing the wheel. I do not mind stealing someone else's program, and I certainly do not mind someone else stealing my program. We are really gearing up, since we have the file pretty well on the way maintenance-wise, to concen- trate on applications; so we are going to try "selling" the file in Columbus now. Mr. Alan Pisarski— I meant to comment on a previous remark. I am sure there will be published proceedings of this conference. So the papers presented here and the discussions will be available. I do not know what the distribution on it or the previous proceedings happens to be. Mr. Silver, could you tell us a little bit about that? As one mechanism for dissem- inating some of this information, I think the proceed- ings themselves are a significant document. Mr. Jacob Silver— All of the people that are present at this conference will be getting a copy of the proceedings. In addition, we provide copies to a number of advisory committees (such as the Small Area Data Advisory Comittee), to all of our coordi- nating agencies, to our census tract key person, plus various professional organizations (such as URISA),to the Washington, D.C. regional and State offices of FHWA, HUD, LEAA, and all of the Department of Commerce offices. That really covers a good number of people and agencies. Just to give you some idea how important and how useful people think these proceed- ings are, we have had the proceedings of each of our five other conferences published and have sold out each time. We now have gone through several reprint- ings of the earlier conferences. Mr. Alan Pisarski— Do the conference proceedings get announced in both the Data User News and other data user publications? Mr. Jacob Silver— Yes, Data User News, through the AIP newsletter, the URISA newsletter, and through the newsletters of several other professional organizations. Mr. Alan Pisarski— Coming back to the question of the internal coordination within a city and how the different elements of the city get together, I had mentioned previously that I was going to ask Mr. Albond and Mr. Tucker to comment on that. Mr. Harvey Albond— Contrary to public expecta- tions, the "grant" is not dead. "Grantsmanship" is very much with us, whether it is in the discretionary funds of the Community Development Act or whether it is in all of the ongoing programs. Even OEO is not dead, the Community Actions Programs are alive and well. Transportation— I chair our regional transportation study— is in the throes of a required and continuous update. What I am trying to say, is that there are various nonprofit social organizations in the public and private sector, for example, the United Way in our particular case, that are looking for economic develop- ment assistance grants. The source of basic data that could be available within the system is the thing. Consequently, if you get the word out at all, you become an invaluable resource, sine qua non, you become something that they cannot do without. 99 In reconciling this public-private thing, which I was wrestling with earlier, I think in listening to our friend from the bank, the fact of the matter is that we are dealing with a consumer product. Government is a series of consumer products and services, and there- fore, the various justifications and measurements we do, whether in the public or private sector, are almost identical— the techniques are almost identical. With a great deal of pride, I can now say that I do think government is ahead of private enterprise, but there is no reason why, again, with appropriate safeguards, and hopefully support, especially at budget time, we can't be of service in that regard also. Mr. Alan Pisarski-I was struck by the thought, after hearing the Director's remarks-you remember that set of curves he put up on the screen demonstrating that costs of a general-purpose survey decline with increased usage-that the Geographic Base (DIME) File, which is obviously not a survey but a general-purpose tool, parallels his point regarding shared costs; and so I think that the same cost relationships hold. We have the same problem with Geographic Base Files; which is, that frequently coordinating institutions do not exist within a city to bring together a set of people to fund and to direct a tool that is of common value to all of them. This is particularly true when some of the players do not even know that the tool exists, or if they did know, it would not impress them unless they could see that tool in action— see its application. Each city has to take its own approach on whether it is going to have a very heavy public dissemination program, or not. But, one of the functions of Federal Government is to assure that the local agency that it is supporting is doing the job of letting the rest of the city know that it exists. Mr. Mitchell Tucker— In spite of that magnificent piece of software that we have developed, some of the data is just downright lousy. This comes from the problem of updating and that is now centered in a city-county agency. That is an institutional problem. They, of course, are having other problems, such as workload problems, which causes them in turn not to contribute all that much. The impact of these newer systems is turning this around, and the Chamber of Commerce in Tulsa has been a big help in emphasizing, or causing a renewed emphasis, in this updating and maintaining of accurate geographic data. The Chamber, of course, understands now that this can be an immensely favorable impact in the private sector, the community. Mr. Alan Pisarski— Would anyone else like to com- ment on the question of dissemination within a city and getting the word out? Mr. Richard Luckay-The W.A. Storing Company is in the private sector. I believe we have the unique distinction of not only contributing financially to the upgrading of the Columbus GBF, but we also have had the opportunity, and the pleasure, to contribute hard data to the updating itself. We provided Mid-Ohio Regional Planning with address and ZIP code data that we use regularly in our direct-mail operations. In effect, we were giving away part of our "family store" when we contributed. But we are interested in this GBF, and in making sure that Columbus has the best GBF in the country. What it amounts to is that data is data and without a viable application it is virtually useless. Once information regarding the GBF— and its potential— is disseminated by means of this convention to the local area and beyond, basic recognition of GBF capabilities will lead to requirements for specific applications. We have many specialized applications that are, today, among the best in the country. But unless the GBF remains in the public domain and is made available for everyone's use as a base data file, our applications will not have a market. Consequently, we are more than happy to be here today and to be able to contribute to the GBF in any way that we can, either financially or data-wise. You might say we are proof-positive that the private sector has the capability to help not only the GBF/DIME update (and the entire program), but conversely to help itself in the process. Mr. Alan Pisarski— I think that is a very pertinent comment. We talked in the past about the notion of a kind of geographic utility, if you will, a public service function of the individual urban area that will provide this general service to the world with perhaps some group like the Post Office providing kind of in-and-out service. Mr. Donald Bufkin— In the Tucson area, the Urban Transportation Study agency got into the creation of a GBF/DIME file for the region very early. As I recall, we assisted by providing local mapping assistance and editing on early versions of the metropo- litan map series and, of course, completed the GBF development and coding prior to the 1970 Census. The prime reason for the early involvement of the transpor- tation study in the Tucson area was that we were the only feasible agency with regional planning responsi- bilities and capabilities. Our region had and still has sometimes sharp and sensitive divisions between city, county, and State planning agencies. 100 The transportation agency initially, and now since late 1973, the Pima Association of Governments (the designated Metropolitan Planning Agency) is the logical agency to function as local coordinator for GBF/DIME and as repository for the information. Since the initial work of the transportation agency in GBF/DIME, the evolving and ever more structured "3C" urban transportation planning process has gotten considerably more confining. May I say for the benefit of the US-DOT people who are here today that you have created the Intermodal Planning Groups and you have created a certification procedure that has forced us to address ourselves more directly to transporta- tions. Our role now in GBF/DIME is more that of coordinating, and attempting to encourage commit- ment and cooperation between city, county and State agencies. That function can be a frustrating and discouraging thing. I wish that I could report that we have a key or that there is a solution waiting somewhere down the way, but the transportation studies that I am familiar with have little staff time to direct their efforts toward GBF/DIME development and maintenance. I have found that the attempts to encourage cooperation among city and county agencies is not bearing the fruit we would like to see it bear in regard to voluntary effort in correcting, updating, and expanding the GBF/DIME. Mr. Alan Pisarski— Those are very appropriate comments. Perhaps not all of that is necessarily accidental. I think there was some conscious feeling in the Department of Transportation that we went through a period, in the mid-60's particularly, where in many cities the only data game in town was the transportation agency. We reached the stage of finan- cial, if not mental, exhaustion. There were some really very bad things done. By that I mean, transportation data sets— which were quite comprehensive, quite sophisticated, and in many cities, as I say, the only set in town— were used for the most awful purposes, with a result that the Department was in a very uncomfort- able position. It comes back again to the Director's point of how do you share the common cost of the independent data set. The way it was being shared was DOT did it and everybody else used it. There was a need, and there was pressure, to create a more balanced, let us say, data collection process. I might mention two programs of the Department that I think would be appropriate to some of the concerns here. The one is the journey-to-work program and that has several facets to it. One, is the tabulation of the journey-to-work data from the 1970 census. That is somewhat aged, but I think it is a good example of what can be done in a kind of local-Federal-Census cooperation. It is to my knowledge, the first time that the Census has been in a position to produce data sets where the requester defined the area system for the data set. What we did was design a set of basic tabulations of the journey-to-work, send those out to the cities and the States for comment, and put that back in for a couple of Iterations, and develope a standard tab set. The Federal government paid for the programing so that these data were available to run, and then a city could define, let us say, traffic zones, school districts, or whatever area system they wanted to use, and it could request the data by that area system. It had to provide the Census Bureau with a block-to-zone or a block-to-area converter. The Census Bureau would then run the summary tabulations up to that zone level, and provide that to the city. The cost to the city was the actual direct cost of the data processing as opposed to any of the start-up costs. I think out of the 200 and some SMSA's these days, about 180 or so have taken part in that program in one fashion or another, with varying degrees of success as a function of the geographic coding. One of the very valuable elements too— because they actually did do origin and destination work between those zones— is that is was the first time that the Census Bureau, in effect, inverted its population file and did tabulations of the work address as opposed to the home address. Thus, a lot of nontransportation people have become interested in the data because it does provide some look at employment statistics. Following that, we are now getting into the melding of journey-to-work data into the Census Housing Survey, which is an annual survey. Twenty cities in the country are selected for fairly intensive sampling each year and that is rotated on a 3-year basis so 60 major cities get covered— 6 cities get 15,000 interviews each year and 15 or 14 get 5,000 interviews as an update of housing and journey-to-work statistics. I think there is another program which will be of some use. A final program I might mention, because it fits very well with what Mr. Campbell was talking about, is something that we call "Special Area Analysis," which is a battery of computer programs that provide the capacity to do a kind of contouring as a function of travel times. It calculates people's accessibility to different services, whether it is accessibility to jobs or accessibility to hospitals. We are talking about the percent of population that can get to a facility in what proportion of time. It leans very heavily on Census data and on network data for the city. Mr. Meyer, I 1! have made several references to the things that I thought were unique. Do you want to comment on them? Mr. Morton Meyer-Mr. Pisarski has touched upon a use of the GBF which is limited only by the imagina- tion of the user. Once the computer has available to it the elements of a mapped network and a way of relating data to the network, much that was formerly impossible becomes practical; and with additions to the GBF file, can even become routine. We plan to add building names to the file for "address matching" purposes and because buildings are also major trip generators. Add building coordinates locally, plus one-way street indicators, plus travel times by street segment during rush and nonrush hours, and the basis for truly sophisticated origin and destination analyses exist. The same techniques apply, of course, to accessibility to hospitals, location of doctors in relation to the population served, etc. The converse is also available, that is, where should facilities be sited to best serve the population of risk. I should mention here that in order to better serve needs such as these, the Census Bureau plans, within the limits imposed by confiden- tiality of individual information, to make available as much summary statistical information as possible for small geographic areas. Mr. Michael Plett— I have a comment on something that has been stressed, but I want to restress it. I am from the greater Cincinnati area, from the Ohio- Kentucky-Indiana Regional Council of Governments, and if our agency had been asked to give a report here, instead of calling it the "Columbus Success," we would have called it the "Cincinnati Failure." The difference between the Columbus project and the Cincinnati project is basically that the Cincinnati project lacked two ingredients that were mentioned yesterday. The first, of course, is cooperation. In Cincinnati, one agency was stuck with the role and nobody was around to really coordinate an extensive geocoding operation. Although we started out as a multiagency, multicounty operation, monies dried up in various agencies, other agencies had no reason to support the continuing effort, and finally only one group was left to carry out the activity— and that was a transportation planning group. The other problem that occurred was that the whole approach was not applications-oriented. People kept asking, "Well, why are we doing this?" and there was no good answer. There was nothing to show the decisionmakers to justify our continued involvement with DIME technology. As happens in all Federally- funded transportation planning activities, as has been mentioned by the gentleman from Tucson, there is a strict set of certification requirement— a, b, and c. At every monthly staff meeting someone would say "Let us spend some time doing some updating on the Geographic Base (DIME) File," and the response would be "What does that have to do with a, b, and c?" Gradually, the commitment of time for the project was reduced. No time was spent on training people to continue this effort, knowledgeable people drifted away, and the project went fallow. We have just spent the past month finding out that there is an application. We are currently required to extensively reappraise our existing (1965 Wilbur Smith) plan, and to adequately achieve this, we must utilize local data. For example, we have 35,000 employer addresses and over a million automobile registrations to geocode. If this geocoding had to be done manually, it could not be done; so, we must have an automated geocoding capability. This implies a need for the most up-to-date DIME files we can get. The trouble is that we have just expended the last month in trying to figure out where we stopped a year and a half ago. The two ingredients that I think cannot be over- emphasized are that you must have a cooperative sustained effort-you need a consortium by a number of agencies, a number of funding sources, and a number of people who can provide constructive input to let you know exactly how to best develop the file— and secondly, you must show initially that you are going to do something with the file. You have got to have applications to begin with; so that you can say "We either develop DIME technology or we are in big trouble." Mr. Alan Pisarski— I think that is really an excellent comment. It focuses on something I said the first day. There are several kinds of applications, several kinds of activities, that really do justify the existence of coding guides and its support, as opposed to having a program that says, "Yes, we will fund the maintenance of coding guides on an annual basis." These specific applications should be the appropriate focus. I wanted to ask you or anyone else to comment on the question of whether there is a logical agency that we can point at in any town that ought to be doing this, or whether, each city should be flexible and try to go wherever the money is or wherever the capability and interest is. Dr. Robert Aangeenbrug— One of the recommenda- tions that I will have to make is do not lean too heavily on higher education. I think in the Cincinnati example, for instance, part of the life that was breathed into the organization was in the hands of one single academi- 102 cian, a very capable one, I am sure, and a very successful one. The first priority in higher education when tenure, promotion, and salary are considered, does not necessarily stress your local involvement in the community. I would very strongly suggest that you beware. This does not mean that these institutions are not an invaluable resource. I think Ohio State, Univer- sity of Washington, and a number of other places have indicated this. But I really feel that unless you can convince the local higher education administration to go into some kind of contractual agreement, you are in a dangerous position. In the case of many institutions of higher education, and I am speaking now as a university planner, currently on leave, as one's prior- ities change in higher education, and they will, local community operations support may be quite vulner- ble. It is one of several tragedies, which you have reported on here that have occurred, but it is a difficult thing. On the other hand, I do not think there is a best individual agency as far as I can tell, and I have meddled or have been involved in at least five separate city activities. In New York, I do not think anyone wants to be in charge. Even the Port of New York Authority with all of its money was not interested, it politically could not cope, I suppose. This, I think, may be a case of diseconomy of zeal in a large organization. I like the idea of the Chamber of Commerce personally, because I think that they are one of the few private agencies that have a community interest at a very high level. I was somewhat intrigued to hear the United Way mentioned. I think we might cultivate that particular cooperative semiprivate-public corporation a little more. I wonder if anybody could react. I really do not think there is a good model. Sometimes a strong individual public administrator will create the same problem that happened many years ago when the GEO-PLANS and related projects were developed with high hopes marketed by a strong individual, with a great personality, fantastic fund raising capacities, who then got job offers elsewhere, and the project went some other place. I simply want to make you aware of that because we are playing with the hopes and aspirations of citizens, and there are a lot of problems out there. They are not getting solved in some of these places, after the prophets have left, so I want you to have a rather juandiced view. Mr. Harvey Albond— There is no doubt that we are in the individual or personality cult. These things are evolving at various levels due almost to the happy coincidence of some leadership, some personalities, some' capability coming together and sitting down to say what is the ideal solution. Looking at our own case study, if you will, we have struggled with the New York State Department of Transportation to decentral- ize the transportation planning process down to the regional level, to have our own study staffs, in cooperation with the State, in order to not only participate in local decisionmaking, but perhaps as importantly, to have some capability of spinning off this data that is evolved for transportation purposes to do planning purposes as "differentiated" from it, and vice versa. It is a feedback system that makes the transportation better because it works with the plan- ning, as such. Now, I would be delighted, frankly, with a limited staff and budget, to dump my process, or shall we say aggregate it upwards, into a regional transportation study process. I happen to be the chairman of the transportation study. On the other hand, I have a problem. My problem is that the money is going to come from UMPTA and from Federal Highway, and so on and so forth, and I am going to have problems getting additional work of a local character back out of them. We are going to have to have a certain degree of duplication. Perhaps we can merge programs and formats, we can allocate our information upward. I think the answer in our case is that we encourage regional models, we encourage State models, we encourage national models; but on the other hand, we are still going to have to have local responsibilities if we want to have local convenience. I guess that we have to pay a price for that. Mr. Alan Pisarski— I would like to shift the focus a little bit and talk about coordination between cities and how the word gets out. I guess I will start off with a general question: Do you feel that you are being served, that you are being sufficiently informed about what other cities are doing? Rather than having a group of people stand up and say "No," I would like you to get up and comment specifically on how that might be resolved, even more specifically, on what might the Federal government do. I am afraid that we might be at that stage in the discussions where we ought to be talking about what should the Federal government be doing? Is the word getting out? Do you feel that you are getting enough information about what is happen- ing in the country? What a) should be done, and b) what should the Feds be doing about it? Mr. Jack Pascoli— You alluded several times to the 15 percent sample of work trip information of the 1970 Census. I am wondering if with 10 years of improvements in hardware, 10 years of improvements in software, and supposedly complete GBF/DIME Files for the 1980 Census, if we can get the work trip information out closer to the Census year instead of 3 years later the next time around. 103 Mr. Alan Pisarski — It is really a good case study. We started designing that process, probably in 1967, I would say. I think the first conference at the National Academy of Science was in 1968, and the purposes were really twofold. The experience from the 1960 Census had been that the Urban Transportation Studies which actively wanted to use the data had ended up in very long queues at the Census Bureau waiting for their turn to do some specialized processing, and at tremen- dous cost. The idea was that by developing a general- ized base set of tabulations, we would reduce the cost, we would reduce the queue time. This was not to say that these tabs would be everything that an individual city might choose to have, and the avenue of special requests at the Census was still open, but simply to speed up this first-round process. Assuming that we were successful, and that 3 years out is successful, I guess if we had gone the old route of the queue and the individual requests, I do not know when we would have gotten through the process. We had, as I say, about 180 or 190 special tabulations run. The total cost now, I guess, is close to a $1 million-around $800,000 or $900,000 spent on the processing outside of the programing— an incredibly expensive operation. If we had gone the individual route it would have probably taken 5 or 6 years and I do not know how much money. It is not to say that it was well done. I think it was poorly done. I think we have learned a great deal and hopefully we will do better in 1980. I guess the Council on the 1980 Census is just being formed now. The focus will be on a better set of journey-to-work statistics, on better geographic coding, particularly the employment end. Mr. Ronald Henderson— This question of how to get cities to communicate, cooperate, how to spread the GBF work, is a very interesting question. It seems to me that there are only two alternatives for the GBF/DIME. One is for the Federal government to require that the GBF/DIME be installed and provide subsequent funding. In a sense they are doing that— the new LEAA directive, HUD requirements in some cases, are a round-about way of requiring GBF/DIME devel- opment. There are a lot of things that are happening, but either the government must require GBF/DIME development, or we will have to go the slow route with trying to talk to local public and private organizations about the advantages and benefits of the file. It is going to take a lot longer that way. Every city is going to have different approaches. There will be a Chamber of Commerce in one city; it will be the city offices in another; the State may be interested in another location. Being of data processing origin and having read Orwell, I often have conflicting opinions on the situation. In a way, to me it would be so much easier if everything were required, and if everyone would have a number, and you had a board with everyone wired so that administrators could watch the board and know who was going to the bathroom; but, on the other hand, we lose quite a bit if we adopt that approach. I do not see any other solution, either it is required or you try to convince people. Obviously, you get more cooperation if you try to convince people of the benefits. I think in a sense, that the government is requiring it in a round-about way. Hopefully the criminal justice people, planners, city officials, and private industry will see the advantages without having it forced down their throats. I think that the approach that is being taken is probably the best one that there is. I think that it should be the responsibility, the goal of everyone, at least here at this conference, to spread the work, to work in their own local area to try to make people realize, so that with what the government is doing and with what private industry is doing, I think we are moving along. Mr. Robert Gebhart— Possibly what I have been thinking of is somewhat simplistic, but what is the possibility of the government agency most interested in promoting the GBF/DIME putting together a monthly or quarterly newsletter which would attempt to collect and then disseminate uses and utilization of the GBF/DIME, actual cases of various people around the country using it. Over a period of time this would give a prospective user who wanted to sell the concept a good documentation of possible uses and examples of people who are actually doing something. It could also promote an interchange of information between users which could help alleviate the problem of reinventing. Mr. Edward Kutevac— As a director of an agency who has been attempting to keep alive local involve- ment in the Geographic Base (DIME) File in our area, it certainly was welcome news that possible there would be some funding available to agencies such as ours. I say this not so much because of the amount of money available but to point out to some of the decisionmakers that somebody knows that you are alive and appreciate the benefits of local involvement. We, in our area, Trumbull-Mahoning County, do not want to reinvent the wheel. It has been the two county planning commissions that have been working within their existing local budgets to supply the various coders and technical personnel to update the base file. I want to commend the city of Columbus and thank them for the aids that they have promised to help sell our Chamber of Commerce and Community Improvements Corporation to the possibility of assisting the local county planning commissions in completing and taking advantage of others experiences in establishing a useable geographic base file. 104 Mr. Alan Pisarski— I think the last two comments with respect to the newsletter and the funding are very appropriate introductions to Mr. Meyer to comment in this discussion on those and on anything else that he would like to toss into the discussion. Mr. Morton Meyer— We know that in the GBF system we have the greatest thing going since the invention of the wheel, and maybe since the discovery of fire. It is true that its adoption to date has been less rapid than we may have wished, but that has also been true of many other innovations. Electronic computers, for example, were strange and forbidding beasts when they first arrived on the scene and only the most innovative agencies or corporations recognized their potential and underestimated the problems of bringing them on line. But today, no matter how much you may bless your computer and its perverseness, it is impossible to get along without one. What I am leading up to is that the demands for spatially-oriented data now being placed upon you— not only by Federal and State governments but, and of equal importance, by the need for local governments to operate more efficiently in an ever more costly environment— lead automatically to the development and maintenance of geographic systems as really the only tool through which the need can be effectively met. By way of further illustration, let me again quote Standard 4.8 of the National Advisory Commission on Criminal Justice Standards and Goals which Dr. Stevens referred to in his comments. It states that "where practical, police should establish a geographical coding system that allows addresses to be located on a coordinate system as a basis for collecting crime incidence statistics by beat, district, census tract, and by other 'zoning' systems such as schools, planning zones and ZIP codes." It further recommends that police should use the same GBF as other city agencies and describes the benefits of using the Census Bureau GBF/DIME (a copy of Geocoding Standard 4.8 appears in appendix 1 following the Question Period of the paper by Ms. Brayford and Messrs. Kelly and Pearson). As regards "spreading the word," we in the Federal system can and do use newsletters, we can and will set up a clearing house, we can and do promote "benefits and uses," but we areJn a sense located on the wrong side of the tracks. That is, we are not part of the local user community. Communication directly between cities is, by far, the most effective way of disseminating GBF/DIME information. For examble, if Mayor Moody, the next time he speaks publicly, can talk about how Columbus is improving its operations through the use of the Geographic Base (DIME) File System, and the importance of the file to a more effective city management, he will do more to promul- gate the use of the system locally than anything that we in the Federal Government can say because he is an elected official, and equally important, because he has the credibility that comes from working with and trying to solve problems that are common to all cities. I would like to conclude my remarks by thanking our host city, Columbus, for its hospitality and again expressing our thanks to Mid-Ohio and Columbus Area Chamber of Commerce without whose yeoman work this conference could not have taken place. I also want to express our sincere appreciation to our distinquished and gracious chairman for his superb direction of the conference during the past 2 days. Summary of Proceedings ROBERT T. AANGEENBRUG INTRODUCTION There have been times at which my behavior at these meetings surprises some people because I usually have a hard time sleeping at these conferences, since they provide one of the rare opportunities to see real live users. Now I am a pusher, as Mr. Barabba knows, of this concept. One of the alleged great-great- grandfathers of DIME was a part-time professor of mine, Mr. William Garrison. I also waddled around, trying to solve New York's problem, and got out, although I got paid as a consultant. It is interesting, I went back last night, being unable to sleep, and since I am now a bureaucrat, could not behave like I did at previous conferences. I actually went to my room and studied. I went back through the proceedings of these various workshops, and found out some interesting things. I think they kind of reflect what is happening, and what has happened, and what is going to happen. First of all, the first one of these conferences was held and planned and schemed by Mr. Jacob Silver, who together with Mr. Pisarski, has a perfect attendance record throughout these conferences. That is, they have attended all of them, as best I can tell from the program. The interesting thing is that these are actually rather active conferences that deal with dissemination of a product that to some extent we still do not quite understand. GBF's are a generic concept, and have been around for a long time. They have never been so exciting in terms of their potential, and we have not had any hits scored until just a few years ago. This particular session is an excellent example of a "con- sciousness-raising" session. The subject is local use of the GBF/DIME System. ary encoding is especially important for area analysis. The term GBF/DIME System implies an operative file system suitable for geographic referencing based on dual independent map encoding. It is our purpose here to see to it that this "system" is of use to local decisionmakers and hence to our citizens. Thus far implementation of such "systems" is limited. Local uses on a regular basis are rare. In other words, we are just getting started. These conferences have docu- mented some of these starts. A quick review from previous conferences indicates that the number and the composition of the attendees reflect not only the increasing awareness of local users to this opportunity, but also finally documented case studies. It is true that we have, except for the Cincinnati case, little documentation of failures. The mix of attendees at this conference shows a greater proportion of local users with the "Feds" less prominent numerically. The dissemination process of this technology still indicates a limited use of these GBF/DIME Systems. The real theme of this conference is about local programs. It is interesting, we only had three new local programs reported, the others had already been reported at previous conferences although there were new update and extension activities. I will give you a brief review of previous conferences as I think there are three things I want to talk about in terms of needs, and these are needs for local and other users. I do not think we can really separate them. First, output— technological needs; second, policy needs; and third, communication needs, which really grow out of both of these. Geographic base files have been and will continue to be in existence for years. A street map with a index is a geographic base file, so are various commercial street directories. DIME is a conceptual model of such a file. More specifically, it is a geographic file structure based on topologic referencing of map features. By itself DIME is an encoding convention of enormous value in point, line, and area referencing. Its feature of bound- SOME REFLECTIONS ON THE GBF/DIME CONFERENCES Since I am currently at the Bureau I felt compelled to take my own census. I compiled a list of attendees at each conference and counted a total of 376. Surrounded as I am in Suitland by tabulators I felt a cross tabulation was in order (see table 1). Not only 105 106 Place Wichita, Kans. Jacksonville, Fla. Arlington, Tex. Seattle, Wash. Boston, Mass. Columbus, Ohio TABLE 1. A Classification of Conference Participants Participants Date Total Local Federal State Research Private November 1970 35 April 1971 41 November 1971 52 January 1973 62 April 1974 88 November 1974 98 Totals 376 11 15 3 5 1 15 16 5 2 3 19 17 7 5 4 20 21 4 6 3 30 30 13 10 5 48 18 10 14 8 151 117 42 42 24 does this reflect the changing mix of attendees, but also the expansion of the number of registrants. Mr. Meyer informed me that many requests for attendance at the current session had to be denied. From a modest size conference of 35 at the beginning, at the current session we have nearly 100 attendees. Using the address list from each of the conference proceedings, I arbi- trarily categorized the participants in five classes: 1) local users— those employed by a metropolitan, muni- cipal, or other local government; 2) Federal users— those employed by Federal agencies; 3) State users— those employed by State agencies; 4) research users— those employed in higher education or by nonprofit research and development institutes; and 5) private sector users— those employed by private consulting companies. Although the largest absolute and relative number of participants are local users, a look at the presenta- tions (table 2) indicates thus far experiences from 26 metropolitan areas have been documented. A further inspection of the table of contents of these proceedings also shows presentations made about several statewide systems and a number of methodological papers by the Geography Division including some critical looks at current problems. I recommend you read these pro- ceedings. I think we had fewer than five women at the first four conferences. Representatives of minorities have been rather scarce, although at this meeting 1 1 women and not so token number of minority attendees are present here today. To me it is a reflection of where we are. One of the big problems we face in this area is the dissemination of technical skills to the nonmajority, and I think that is a reflection, somewhat personal on my part, that I find rather interesting. Maybe you will be able to do something about it because, let me tell you, if you complete an academic program that combines geography, data processing, and do not ever forget, at least a course in calculus and statistics, you could probably find a job some place no matter how hard times get economically. I think there's some interesting opportunity areas for affirmative action here. TECHNOLOGICAL NEED IN LOCAL AREAS The specific technical requirements of a GBF/DIME System are not widely understood by most data processing managers or city planners. Perhaps they should not be, but one may ask— Is there in each SMSA one key person who could conduct a 2-hour discussion of this system for, say, the planning and data process- ing staffs in various metropolitan agencies? Could this person identify the specific hardware, software, person- nel, and monetary requirements? And, could he or she make a case before the budget officers that would convince them of the utility of this system? I would venture to say in only one-quarter (about 60) of the metropolitan areas could such a person be found to give such a speech, and in only one-eighth of the SMSA's might he or she convince someone to provide additional personnel for control of files, cartographic programs, and statistical analysis. URBAN BASE MAPS The base map problem is quite often overstated. People vent their spleen against the USGS (United States Geological Survey), Mr. Meyer, Mr. Fay, and whoever else they can find. It is often an irrelevant argument. If you understand the local scene, you can in some crude form develop a local map. If you cannot, you do not belong in your current job. You do not even understand your own environment, and you have no capacity to fail. (Or, to put it in other words, you don't want to assume responsibility to do it yourself .) If you have no capacity to fail you do not belong in the public sector, or for that matter, in the private 107 TABLE 2. Metropolitan Areas Documented Conference site Wichita, Kans. Date 1970 Jacksonville, Fla. 1971 Arlington, Tex. 1971 Seattle, Wash. 1973 Metropolitan areas documented Wichita, Kans. Washington, D.C. Orange County, Calif. Atlanta, Ga. Kansas City, Mo. (see also Arlington, 1971) Jacksonville, Fla. Charlotte, N.C. Dallas, Tex. (see also Arlington, 1971) Minneapolis-St. Paul, Minn, (see also Boston, 1974) Tulsa, Okla. (see also Columbus, 1974) Cincinnati, Ohio Pittsburgh, Pa. Wichita Falls, Tex. New York, N. Y. Santa Clara County (see also Wichita, 1970) Kansas City, Mo. Dallas, Tex. (see also Jacksonville, 1971) Seattle, Wash. St. Louis, Mo. Albuquerque, N.Mex. Miami, Fla. Boston, Mass. 1974 Columbus, Ohio 1974 Las Vegas, Nev. Columbus, Ohio New Orleans, La. Cleveland, Ohio Minneapolis-St. Paul, Minn, (see also Jacksonville, 1971) Columbus, Ohio (see also Boston, 1974) Fort Wayne, Ind. Indianapolis, Ind. Niagara Falls, N. Y. Tulsa, Okla. (see also Jacksonville, 1971 ) sector either. I think the base map question is a serious one, with some serious academic and other problems related to it, but I think it is sometimes overstated. It is on the other hand, also understated. If you have a file that has technical requirements for say, land record- keeping, the accuracy of the file becomes important. A case, that I can remember in southern California required large scale accuracy for say, water systems, fire hydrants, supporting the public safety system. You will have to kind of define the accuracy needs for each problem, some of these scale requirements are very detailed. In this southern California city it was discovered that its map had not been corrected after the earthquake. They had some fairly serious problems with this. I would hope that we can address ourselves at some technical conference not too long from now, to this question of accurate base maps. It is also a national problem, but I do not think it's fair, nor proper, to expect the Bureau of the Census to solve this problem by itself. Perhaps at a later time, we will be allowed to report on some of the Federal cooper- ative efforts that the Bureau has been very much involved in that address this area also from a policy point of view. There are some other problems. However, many of these technical questions can be solved. I think as Mr. Pierce has shown in Albuquerque some time ago, if you 108 want it bad enough, you can get rather good and accurate maps by the use of current technology, and you do not need remote sensing necessarily. Ortho- photoquads and current technology will allow you to do it. All you need is the will and the money which can be raised by creating broadbased State and local support. For each metropolitan area one agency should be able to provide a single base map series which can support the local and Bureau needs for the GBF/DIME System. Whether this is the case or not is a political rather than a technical problem. The problem of keeping a map current is also a management rather than a technical problem. SOFTWARE One of the other problem areas is that of software, and it is a serious one. Again, I find it rather surprising that at this particular conference, not too much public discussion about the software took place. I think part of this is also based on your own professionalism. With this kind of a mixed group you are somewhat hesitant to publicly address possibly the wrong audience. I did find a whole lot of informal communication going on about software problems. I would certainly urge that special sessions on specific software problems be held at future meetings. Just looking back through some of the documenta- tion, we are still talking about some software being developed that needs 256K bytes to run. We are talking about large programs. Some of these are hard to maintain in the current environment. Yes, they are technologically possible, but they are not always locally practical. I think we must try at some future time, and I hope very soon, to address this question. Because if the Bureau makes a decision to develop a BPR type package (developed by the Highway Ad- ministration for transportation planning and analysis) which is an excellent example of a highly transferable complex system; the Geography Division, which would certainly be involved, would very much need your input. The Census Bureau needs very much to structure such a program based on its own as well as local. State, and Federal agency requirements. Aside from the question of making such a package IBM compatible, there are some other questions. For example, what type of an operational environment can be assumed or, should be supported. By 1980 we do have some serious problems in terms of what kind of languages will be used. I know from a practical point of view, we will probably stick with COBOL and FORTRAN, but I am not really sure it is wise. After all, we are now talking about manipulating potentially large files with extremely complex geo- metry and geography. I am not sure it is wise for us to think about general-purpose COBOL programs. On the other hand, I do not know what the solution is. As Mr. Meyer says, and he is quite proper and I do agree with him, the basic requirements of Census geography are not necessarily those of your local environment. There is no way in which the census can serve all with endless capacity. I do not think that is efficient. I do not think that it is proper. At the risk of preaching I would like to list some needs of local users which should be addressed at the next conference and hopefully documented in reports: 1. Document uses of each software module in the GBF/DIME System & CUE, for each SMSA; iden- tify each agency, personnel, and hardware environ- ment; include estimates of costs. 2. Develop software for use of GBF/DIME in nonmail (nonstreet address) areas. 3. Provide a test module for interactive graphic editing (small, documented, and transferable). 4. Provide additional special-purpose packages (i.e. CRAM, CARPOL) for intersection analysis, point- in-polygon search, and selected polygon equiva- lency boundary files. 5. Urge the Census Bureau to hold a technical confer- ence for advanced users on, for example, address matching problems, computer graphics, small-area statistical mapping, and utilization of imperfect and GBF/DIME-Files. QUALITY CONTROL QUESTIONS I have not yet seen sufficient evaluation of quality control reported about some of these files. I think some of us are beginning to get nervous about 1980, or should I say 1979, or 1978, when the final decisions are made. My current information shows we will have the same things we had in 1970— "You like it? You used it?" There are some rather serious problems as far as I am concerned. It would be useful to know if and when quality control evaluation will be done and what the results of the Bureau's research and development efforts and plans are. I think that the people in the front line, particularly the Geography Division, will have an extremely difficult time answering all those questions unless they stand in the larger body politic of 109 Federal programs and the Bureau's overall programs. Several specific questions highlight my concern: 1. How do we define metro-map accuracy? 2. What are the standards of quality for street-address files? 3. What is an acceptableGBF/DIME-File?What are the objective criteria? 4. What is a match (as in match rate in GBF/DIME System or UN I MATCH)? 5. What is an acceptable match rate per block, per tract, per street? URBAN CARTOGRAPHY AND SMALL-AREA ANALYSIS There is another rather vexing problem that has been glossed over thus far, and that's the question of cartographic process; the danger of a new whiz-bang approach. That reminds me of GEO-PLANS and the slides we had of the three-dimensional map of the population of Wichita. In color yet, rotating popula- tion densities through the computer. It was very exciting. We did not know what it meant, necessarily, but it looked pretty and people thought it was great. It was the only map that we could show to the governor and the mayor, and God knows who else, but its meaning was not very well understood. With the advent of improved communication we can tell messages that may not be accurate and cannot always be proper. I have some concerns about, for instance, what I referred to yesterday— the problem of ecological correlation. It is not new, but it has not recently been addressed. I have not seen any recent contract research that addresses this issue. We could indeed portray in map form some enormous misunderstandings among our policymakers. I think if you go back to Firey's work in Boston a long time ago, one becomes aware of the danger of making the wrong decisions based on that kind of communication. We know very little about small-area data analysis. At the Bureau, sensitivity to confidentiality and a concern for statistical standards are well documented, but little or no documentation about small-area data problems is available. For example, the variance of a sample increases with a decreasing sample size. This implies that at the tract and especially at the block level our statistical estimates are subject to a large relative variance. Although there is a great demand for transferring current population estimation method- ology to small-area forecasts, one can easily appreciate the difficulties of such estimates. In addition, the degrees of freedom available for using regression techniques further limits forecasts for very small areas. But with recent legislation the demand for detailed area data, including forecasts, grows. I have asked recently in the Bureau who uses small-area data and what do they need, and to be perfectly honest with you, I could not get a complete answer. And answers are needed to plan for the 1980 Census. These are some real tough questions, only some of which are statistical, and I think the Bureau is probably one of the places I know that can possibly answer such questions. It is, by the way, an important issue particularly with rapid communication, including, say, cable TV. In even a small city like Lawrence, Kans., where I am very much involved, we have local communications through the cable TV, and at such meetings we can show results of some of our statistical analysis of small-area data. To be perfectly frank, we have not always examined such presentations carefully. I would urge the Bureau and other Federal agencies to examine these questions. Now, do not think the Bureau is not sensitive to the problems of small-area data. I am simply saying I would want you to be involved in part of this dialogue also. As a matter of fact, I think the Bureau is probably the one Federal agency that is most sensitive about data use, with the exception of the journey-to-work information where I think DOT (Dept. of Transportation) has historically led. POLICY CONSIDERATIONS We are all here supposedly to improve the citizens' lot, and in so doing, keep the cost of local government down, not necessarily decrease it— I am not naive, that may not be possible— but keep it from growing faster than it is. We need proofs of utility, we need more mayors, elected officials, and senior level managers at these conferences. We need to prove that we have solved some problems. As one of my favorite city managers said, "I will never take this to the council unless I can tell them that I have saved them some money that they would not have saved otherwise." Armed with an improved taxation and evaluation in the assessment file created for his city, he convinced the council that they would be collecting more taxes, more equitably— it sold. I think these are the kinds of proof we need. Case studies could document some of these experiences. I think we need to encourage public administrators outside of the Bureau to be prepared to present results of problems solved through GBF/DIME System technology to elected officials and to the voters. 110 The ownership of the files also raises some very important questions. One which we discussed and discovered in the USAC project cities is the right of public inquiry. Please do not whip up too much emotion about the greatest municipal information systems. I have been there. The citizens may expect you to deliver. The other problem is, are you prepared as a manager to give them the information they desire? Not only are there legal questions but other political questions. For example, would you want to slow down a city government's computer by allowing time-sharing inquiry of the status of all citizens' municipal utility bills, or all the land use records, or whatever else a citizen would like to know? These are rather serious problems. I hope that in any systems design work you do, you involve where you can, people who have policy experience and responsibility. I would urge you to particularly read up on this if you can. I think Public Administration Review and a few other journals, for instance, may be able to provide you with this. Another aspect of file ownership is, of course, the question of regional versus local and the question of consolidation of public information systems. think we must as professionals see to it that the academes deliver us understandable and better models. It will be much better than getting them involved in taking over GBF/DIME-File maintenance relationships within urban areas. It is not really their responsibility, as far as I am concerned. COMMUNICATION OF GBF/DIME TECHNOLOGY Obviously this kind of conference provides excellent opportunities for face-to-face and formal communica- tion. But what happens tomorrow? I would like to briefly touch on some opportunities available to you. In terms of local-to-local communication, URISA (The Urban and Regional Information Systems Association) is an organization that has a SIG-GEO organization, that means a special interest group devoted to geographic base files. For $2 per year you can join and receive a newsletter. Mr. Richard Schweitzer of the Geography Division is chairman of this group. You can write to him at the Bureau. An important dimension of new public technology deals with the "credits" of its results. Usually the chief excutive of any corporation owns anything he or she likes; I think that is only proper. When you are using this new technology you get into the serious question of having an extremely powerful tool which could lead to an interesting tug of war as to who is going to take the credit for it. Now, I am not telling you very much new, but I do think and I want, in this conciousness- raising session, to leave you with the impression that you should try to push the credit as far down and up as you possibly can. I find that the most impressive managers I have seen are those who have their staff take a lot of the credit, very far down in the organization and inform their superiors. I think if you talk about regional cooperative models, having some- body else take the credit is an excellent strategy. Another real serious policy problem is the lack of adequate theory and models. We may be creating some kind of a monster. We do not really understand how local economies work. Nor do we understand some of our medical and other types of problems, including welfare and related phenomena. We do need some more theory. It is not really your business, per se, to deliver that. Most of you are from operating environ- ments where you do not have time to deal with these, but decisionmakers need to have these kinds of models. They have them or invent them, you know. They may be such models as "All those folks over there always try to give me trouble, but they do not have much clout, so I'm able to ignore them." This is a model. I I was afraid I was going to have an argument with some of the Federal bureaucrats who were going to develop some stupendous proposals for more Federal dollars to provide communication among the locals. I really like what I heard from Mr. Meyer. Such communication must probably be informal and you the user should control it. It's not that expensive to do. I think Mr. Barb and Mr. Cooke, who preceded Mr. Schweitzer in this SIG, kind of invented the use of NTIS (National Technical Information Service) for disseminating proceedings of the annual URISA Conference. I would urge you to attend some of the annual meetings or special workshops of URISA. Additional geographic interest groups are found in ASPO (American Society of Planning Officials) and AIP (American Institute of Planners). Another way of expanding the utility of the GBF/DIME System technology is to attend the meetings and/or workshops sponsored by the Census Bureau. You might also consider pushing the tech- nology by preparing your supervisors to read papers at their professional societies. So what if they steal your thunder. If you provide the script they are bound to be interested in supporting the GBF/DIME System tech- nology. At the international level there is a great deal of interest in this technology. The URISA newsletter and the International Geographical Union can provide you information about your foreign colleagues' ex- periences. 111 Communication is a problem, and local communica- tion is no exception. I will not belabor it because most of you probably understand it better than I do, but it certainly involves problems of cooperation, coordina- tion, and again the cult of personality. Learn how to eat humble pie, don't get in the way and take too much credit. Get your subordinates or your superiors to take the credit for what you have done. You will enjoy it. I think it is one of the opportunities we have missed in some of the experiences I have watched. In parting I have one more suggestion. The geo- graphic base file technology is ultimately a local tool. It is a new tool, somewhat complex, but an ideal index to analyzing and monitoring local and regional phe- nomena. Once the standards for geographic informa- tion systems are set, the dependence on local update and maintenance will produce a "live map" for each major metropolitan area. At future conferences we, the "Feds," will be learning more from you. Library of Congress Cataloging in Publication Data Main entry under title: Geographic base (DIME) system— a local program; conference proceedings, November 18 and 19, 1974, Columbus, Ohio. (Computerized geographic coding series GE 60; no. 6) 1. Geographical location codes— Congresses. I. United States. Bureau of the Census. II. Series. G108.7.G47 001.6'424 75-619097 U.S. GOVERNMENT PRINTING OFFICE SUPERINTENDENT OF DOCUMENTS Washington, D.C. 20402 OFFICIAL BUSINESS SPECIAL FOURTH-CLASS RATE BOOK PENN STATE UNIVERSITY LIBRARIES POST U.S. DEPARTI\ ADODOTEflS^?! u.o.rviMIL •OV-UT/Oa,