Interagency Cooperation
in Information Management:
Why, Why Not, and some ideas on How*
Jeff Waldon, Assistant Director
Conservation Management Institute
College of Natural Resources
Virginia Tech
Without the information to make good decisions, rational management of natural resources is not possible. Acquiring, managing, and distributing appropriate information for biodiversity and land management issues is an expensive and time-consuming job. Although the research capability in the United States is producing more information than ever, presumably for managers, our ability to synthesize and apply that information has only recently advanced to a stage where reasonably useful systems are available to access the information in a useful timeframe. Although over 100 peer-reviewed journals are published, and innumerable gray literature reports are released, and the World-Wide-Web has blossomed to a point where at least an order of magnitude increase can be expected in the availability of fish and wildlife information, it is all a pointless exercise in paper shuffling if managers can't use the information to conserve and manage our natural resources. The current explosion of information has been likened to "getting a drink from a firehose".
Why?
Paradoxically, information on fish and wildlife is still fairly sparse in many important areas. Particularly, distribution, abundance, and basic life history data are missing for the majority of vertebrates and almost all invertebrates in the United States. This also happens to be the most expensive and time-dependent information with which we deal at a time when budgets for survey and monitoring of biodiversity in general are shrinking. Data sharing that reduces duplication of effort can greatly increase the quality and quantity of information available to managers. This activity reduces cost, improves decisions, helps to better satisfy the public, and in general gives the data sharers more credibility as knowledgeable professionals.
Why Not?
Although data sharing (along with Motherhood and Apple Pie) is generally considered to be a good thing by agencies, surprisingly little takes place. The usual list of problems encountered include:
- Proprietary attitudes towards data by data generators.
In many cases, data collection and management is directly related to the career goals or budget of those holding the data. There are few direct incentives for data generators to freely share data with other agencies or individuals and some important disincentives. Data generators are typically funded to collect data but seldom funded to distribute it. In some cases, distribution is an important source of benefits for the data generators in the form of funding, publications, or political power, and they have a powerful interest in restricting access to the data to maximize its value. In other cases, agency budgets are linked to the ability of the agency to provide "answers" of some sort to public policy questions such as permit reviews, plans, and decisionmaker inquiries. Having the data to address those needs, justifies budgets and jobs.
- Inconsistent data definitions.
Unique data definitions further set apart agencies, professional societies, and even universities. Again there are few incentives for developing and adopting standards and several important disincentives. Using data definitions that set agencies' data apart, makes the dataset unique and less likely to be reduced in subsequent budgets because of its uniqueness.
- Lack of time and/or trained personnel to perform data transfers.
This is a real problem in many agencies. Data collection and sometimes management tend to be higher priorities for agencies than data distribution. Techniques for data sharing require a higher-level of training and knowledge of other available datasets than simply capturing and managing a single dataset.
- Lack of knowledge regarding other potentially useful datasets.
Although indexing of datasets has been ongoing for decades, the number of individuals and programs collecting data seems to be ever increasing. Some very good indexing efforts are underway, and yet the scope of the problem is immense. For instance, of the thousands of scholarly articles published each year on some aspect of biodiversity, presumably a large proportion are based on data. Dataset indexing is probably lagging behind by an order of magnitude at least, and of course this is additive over many years. In some cases, there are no incentives for acquiring and using additional data, so no effort is made to see if additional datasets are available.
- Lack of knowledge regarding the potential usefulness of datasets to others.
In many cases, datasets are collected and later destroyed because the data generators don't understand the value of the data to management agencies or other researchers. This can take the form of a damaged computer disk, a fire, or an overaggressive spring cleaning of the local university store room. Long-term archiving of data is a foreign concept to many organizations and agencies.
- Actual and/or perceived legal restrictions to data distribution.
Some agencies and organizations restrict data sharing citing legal restrictions for sensitive data, fuzzy licensing agreements with the original data source, or concerns about liability. In some cases, these are valid concerns and in others they are not. In almost every case, some form of the data can be shared that meets existing legal requirements either by reducing the precision, passing along some form of data license, or using a liability release agreement. Restricting access to publicly funded data, should be a carefully decided course of action. When data accuracy and adequacy for a given application, cannot be verified independently, the public can justifiably question the public policy of funding such data collection activities.
(*--Note that all of the above problems are policy- or personnel-related, not technical in nature.)
How?
Cooperation, coordination, and communication should be the mantra of all data managers in fish, wildlife, and land management agencies. The most recent manifestations of these include ecosystem management, partnerships, and a renewed interest in data standards. There are two practical rules to help agencies that data managers have used successfully over the years when interagency cooperation (such as data sharing) is the goal.
- Understand the needs of potential cooperators.
Each data generator and manager works in a context of laws, policies, personalities, budget, and time constraints. The mission of the agency or group and the goals of the individual must be taken into account to successfully develop a mutually advantageous relationship. If benefits to all parties are not obvious and equitable, the cooperative initiative is doomed. The artistry of developing cooperative relationships comes in the identification of potential benefits that can be incorporated into the cooperative activity.
- Communicate
The single most likely reason for failure, is a lack of communication among the parties. Although this seems obvious, it can be difficult and expensive, and is often neglected. Meetings, memos, newsletters, and phone calls all take time and resources, but without good communication, success will be unlikely.
In my experience, these two rules, along with the application of basic professionalism and civility, set apart the successful program that easily initiates cooperative activities including data sharing. The adoption of common data standards, regular data transfers, mutually supporting data collection, and common reporting/distribution formats are all technically possible. At present, agency incentive systems are not entirely supportive of such activities, but if current trends hold, that could change. Reductions in agency budgets and fierce competition for declining funds should encourage data collectors to work more closely together. As data collection programs are cut, more attention will be given to reducing overlapping data collection programs, and getting more "bang for the buck" out of existing programs.