and the State government and the local government. These interchanges are shown by means of a code on the chart.

A white square indicates essentially no information exchange. Where a gray square appears, we found information exchanged on an order of approximately 1,000 documents per day; the black squares are those instances where more than 1,000 documents are exchanged daily.

The small corner shown here is a magnified section of this chart. It shows a particular information exchange; namely, the information which flows from counties to the department of public health. In this case the exchanged information will supply data to public health in the areas of vital statistics, diseases, and air sanitation.

In sum total we logged 1,015 interactions between organizations. This, incidentally, is information exchanged between organizations both at the State level and the local government level.

A great many more of these heavy exchanges exist, but also significant was the fact that we found so many white squares. Indeed we discovered organizations which wish they could get information from such-and-such an agency, but discovered it is not available.

One example here concerns hospital patient characteristics-information which the State department of public health would like to get as routine business from hospitals, but which is currently not available, except as an expensive one-time research project.

To give you an example of the extent of information exchanged between local government and State government, we prepared this chart, figure 8, in which, in the center, we have listed such State

organizations as agriculture, education, and so on, which typically concern themselves with information from local government.

In the outside panels we have listed typical county organizations such as the assessor, tax collector, and sheriff. We have drawn lines between the county organizations and the State organizations, the thicknesses of which are semiquantitative depictions of the amount of information flow. Where there are thin lines there is a relatively small amount of information flow; the thick lines indicate more information flow.

The circled numbers indicate the number of reports which are routinely submitted annually between the county organizations and the State organizations. You see 27 from the controller, 204 reports to social welfare, and so on. In the aggregate there are approximately 10,000 reports per year per county submitted from local government to State organizations.

Turning aside from the information flow patterns to the statement of the problem, we find that there are several major information needs in California; we have categorized these in the four sections shown in this chart, figure 9.

In the first we found the need for a common handling of information exchange, a common identifier, a number of which would uniquely identify geographical location to the county assessor or to the State water resources people or to the welfare department. Such common identifiers are needed in the other areas of utilization.

We also found the need for an information locator. Many organizations told us of collecting data which they suspect is also collected by someone else. Without knowledge of its location, we duplicate that collection function unknowingly.

We saw the need for a single data entry in which an information or data item reaches a particular organization, say justice or welfare, but at the same time is transmitted simultaneously to other organizations that have a need for the same precise data item.

And finally we saw, as the State organizations become more computerized, a situation where the present inadequate mailing system would be replaced by an appropriate data communication system between the computation centers.

With these needs in mind, what would be an appropriate strategy for the State of California to follow in the development of an information system? We certainly don't want to go out and suggest that we duplicate the computation centers and the computation equipment which is being implemented throughout county levels, city levels, and State levels in special districts. This wouldn't make any sense.

What we would like to do is to find a way to meld these into a system which would provide a total information system, but would still meet all purely local requirements. We identify this by means of a question mark, figure 10.

The problem, of course, is that while cities and counties are implementing their own data systems, they are moving in different direc

tions to serve their own strictly local needs. We would like to guide them in one direction by means of the items which are indicated by the question mark.

What are these items?

No. 1 is the statewide information data work, figure 11, which would link together the electronic computation centers throughout the State, whether they be on the State, county, or city level. Such organizations as the department of justice, which requires data from counties and cities, could obtain this information on a regular basis by querying directly the indicated files. Similarly the organizations could periodically send out an all-points bulletin which would query all the organizational computer centers of other jurisdictions for needed information.

Thus we visualize a system which would operate like a statewide telephone network, except the customers or telephones would be the computation centers now being created.

The second element of our original question mark involves the creation of a State information index, figure 12. This would be a computerized information file located at some central point in the State and containing the location of all available information. We don't propose that the State go out and create a master State data bank, but we do propose that it create an index which would have the same effect as a data bank, except it would eliminate the agony of geographical centralization of all the files in the State.