Index

 

State Innovation Strategy

 

State governments would need considerable empirical knowledge to innovate successfully in their difficult tasks. For each of these tasks competing factions would propose conflicting objectives as estimates of the common weal. And experts would have numerous conflicting policy proposals to achieve each of the various objectives. Hence, the ability of the state governments to innovate depends on their ability to accurately predict the consequences of selecting a particular objective and associated policy. Because the social sciences are likely to be characterized by numerous controversies for the foreseeable future, accurate knowledge of the consequences of selecting an objective or associated policy would require an implementation of that selection.

States would obtain some empirical knowledge of their alternatives from two sources. First, variations in the objectives and policies implemented by the various states would provide useful empirical knowledge. Second, the federal government in its research and development role would perform experiments and pilot studies. In this section we shall discuss how the proposed design promotes state government innovation by increasing empirical learning.

An important consequence of the proposed decentralization would be more variation  in the estimates of the common weal. Consider, for example, managing the risks of drugs. Granted, it is efficient for the federal government to test new drugs and determine their side effects; however, having the federal government set a single nationwide standard for risk greatly limits empirical knowledge of the correct balance of the two types of risk in drug release into the marketplace. The first of these is the risk from some unforeseen, dangerous side effect. The greater the testing of a drug the further this risk can be reduced. The second risk is that the longer the testing period, the more individuals would be denied the positive effects of an effective drug. Currently individuals journey to foreign countries in order to obtain the benefits of such unreleased drugs, and critics claim that the federal Pure Food and Drug Administration errs by placing too much emphasis on reducing the first type of risk while paying too little heed to the second type. The estimate of what balance between these two risks is in the common weal is a judgmental, not an analytical decision. Therefore the risk management policy of the appropriate amount of testing would be decentralized to the state in order to obtain a better long run estimate.

In the proposed drug testing and release program, the federal Pure Food and Drug Administration would test the drugs and create and maintain databases with complete descriptions of the side effects of drugs and status of their testing. These databases would then be accessible to the individual states and all interested parties. The more rapidly states released promising drugs, the greater the state would need a mechanism to warn potential users of side effects as they are discovered. States would implement their drug release policy by using the data support system maintained by the federal Pure Food and Drug Administration. Most states would probably establish a drug administration to establish a release policy based on the amount of testing completed by the federal government. This release policy would be integrated with the decision support system used by doctors to prescribe drugs. Expert system advances in medicine will eventually create programs which will select the best drug for a patient's illness conditional on his past medical history. A state could integrate the drug release policy into these systems so that the physician would be informed, for instance, that a drug was available without restrictions, required informed consent to the specified dangers, or required consent of the drug release agency.

In formulating its drug release policy, each state would have to estimate the socially desirable balance between the two types of risks. Given the forecasted differences among the reorganized states, state estimates of a desirable balance between the two types of risk would vary considerably. Part of the federal government's role of research and development in innovation would be to have the empirical consequences of these variations systematically analyzed. Over time through this empirical knowledge a better estimate of the common weal in the balance between the two two risks in drug release would emerge.

Another governmental task for which variations in the estimates of the common weal would be very beneficial is the government's role in providing medical services. Medical research has created medical procedures which are effective but which are so expensive that providing all individuals unlimited access would be extraordinarily expensive. Variations in both the level of care to which individuals are entitled and the methods of providing medical services would be essential for innovations in the production of medical services.

A second important consequence in the proposed decentralization is that alternative policies to achieve social goals would be implemented simultaneously  rather than sequentially. Consider the problem of selecting effective policies to manage the environment. Under the proposed decentralization of environmental management, the federal government would be responsible for environmental research and environmental coordination among states by specifying air and water quality at state and federal boundaries. With this coordinating power, the federal government would control such problems as acid rain and pollution of coastal waters. States would thus have operational control over how these federal boundary standards were to be achieved and, moreover, would assume full control over local pollution.

Giving states full control over local pollution increases the likelihood that alternatives policies would likely be tested simultaneously rather than sequentially. To illustrate why decentralization would lead to a much more rapid testing of alternatives consider the evolution of federal environmental policy from 1970 to today. There are a wide variety of administrative and incentive approaches to controlling pollution. One administrative approach is technological-based emission standards and examples of incentive approaches are emission taxes and market incentives. One proposed market approach is the concept of pollution rights. For example, given an environmental target such as so many tons of a particular pollutant released per year in a particular metropolitan area, firms would bid for the right to emit a percent of the target pollution level for a locality. The purchaser could subsequently sell this right.

The federal clean air and water legislation of 1970 and 1972 mandated the states achieve the federally designated standards using only one approach, technological-based emission standards. The states were given little discretion in setting the standards or in devising alternative methods of achieving the standards. In 1977 amendments were made to make this legislation more effective. Now in the 1990s under the Bush administration environmental risk management is finally moving to test the effectiveness of implementing pollution rights, a market incentive approach.

The centralized approach to environmental control illustrates that as long as two decades can ensue before the alternative approaches are empirically tested. However, under decentralization of environmental risk management these alternatives would probably have been tested simultaneously. Environmentalists generally dislike the market approach out of fear that once pollution rights are established they will be hard to adjust. Thus liberal, environmentally oriented states would probably have opted for technological-based emission standards. In contrast, conservative, market oriented states would probably have opted initially for market incentive approaches to pollution control.

Nevertheless, decentralization in itself is insufficient for there to be variations in the policies among the states. For example, if the political majority of each state had the same views there might be no variation whatsoever. However, the reorganization of the states each decade and greater freedom of location made possible by technological advances would accentuate the differences in the political economies of the states. Thus in an political environment where political factions dispute ends and experts dispute means decentralization of tasks to the states would result in considerable variation in both ends and means. The empirical consequences of such variations provide the empirical knowledge for innovation as states would imitate successes and abandon failures.

The systematic analysis  of these variations in state goals and policies would be part of the federal government's research and development task in state government innovation. Scientific information policy and federal research grants would provide the observations and financial support for this systematic analysis. The federal government would also fund experiments and pilot studies advancing state innovation. Because of cost considerations most of these experiments would be conducted on much smaller institutions than the states.

Consider, for example, medicine which would become one of the more difficult problems for the states. Most federal research in medicine, up to now, has been directed developing effective medical procedures without any consideration of the cost. The public has come to expect that these procedures should be generally available to all patients. In informational society the federal government would devote considerably more research effort at alternative efficient methods of medicine. For example, the federal government would fund research to investigate the positive effects of various type of preventative medicine. Given the cost considerations most experiments would likely be performed on towns not the states.

In the area of risk management the thrust of the federal government's research and development efforts would be to create incentives for the participants to reduce the risks themselves. These incentive systems would be based on operational information policy which in itself would create incentives by making all the participants in a risky process better informed of their respective risks. For example, consider decentralized environmental risk management. As technology advances the cost of monitoring the environment in great detail will fall. The federal government part of its task in providing voters with detailed comparisons in subordinate government performance would provide residents detailed profiles of pollution in real time in their homes, workplaces and local environments. Subordinate governments and private parties would make additional measurements. Already, the concept of an environmental right-to-know is moving information policy in this direction.

Consequently, because voters would be much more aware of environmental hazards, voter pressure to enforce environmental laws would be continual instead of varying with environmental catastrophes as is currently the case. In addition, private parties and local governments would pursue a much greater number of civil suits to enforce pollution standards. Faced with greater public awareness and a greater probability of a civil suit, many firms would seek the positive public relations of maintaining a good pollution record. Finally, the states would have much stronger incentives to constantly innovate in environmental risk management.

However, some types of risk management would require extensive research in order to create better incentives. Consider worker's safety. The original worker safety risk management was common law. In the progressive period states created workman's compensation programs which provide workers immediate compensation for injuries without having to pursue a civil suit. In the 1960s the federal government added the Occupational Safety and Health Administration, OSHA, to establish and enforce worker safety standards. Under the proposed decentralization the federal role would be reduced to research and development of safety programs and OSHA's role of establishing and enforcing safety standards would be delegated to the states.

To forecast how the federal government research and development task would promote greater innovations in safety, consider the limitations of the current worker safety system. The current worker safety system creates few incentives for the participants themselves to promote safety. The small number of firms for which safety is regulated by common law have economic incentives to avoid negligence in worker safety. Large firms have economic incentives to promote safety in their firms to qualify for reductions in their workman's compensation insurance premiums.

Indeed, under the current worker safety system most of the incentives are perverse. Since workers do not pay for workman's compensation insurance they have no incentives to promote safety in order to reduce the insurance rates. And in hard times workers have incentives to file dubious claims. Firms have incentives to lobby state governments to the reduce the workman's compensation benefits to workers and to lobby the federal government to reduce the number of OSHA inspections and the magnitude of the fines for safety violations.

In informational society states would have the difficult task of deciding whether to promote safety through establishing and enforcing worker safety standards or trying to create better incentives such as by empowering employee committees with the right to act to monitor and promote safety at their workplaces. As advances in the technology would create opportunities to design much better incentives systems, many states would opt for this approach. And the federal government in its research and development role would fund pilot projects.

For example, consider how advances in the social nervous system would enable the insurance-rate-reduction incentives to be extended from large firms to small firms. The problem is the size to the sample required to demonstrate a better safety record. Considered individually small firms are too small to establish what variations in their safety program are statistical significance in improving worker safety; therefore insurance firms have no valid procedure for granting insurance rate reductions to small firms who have superior safety programs.

The type of pilot project to establish a valid statistical procedure for granting insurance rate premium reductions to small firms would be to have the trade association in conjunction with the federal research and development agency, state insurance board and insurance industry representatives to systematically test safety alternatives across the firms in the industry. Consider just one example worthy of testing. Currently rather than test workers for drugs some firms are testing workers each day for eye hand coordination. The pilot project would systematically test the use of such a machine across the industry. If the use of such a machine were statistically demonstrated to reduce accidents then its use would qualify firms for a rate reduction. The pilot project would also experiment with various combinations of splitting insurance premiums between firms and workers. Thus through systematic safety experimentation throughout an industry insurance rates could be based on installed safety equipment, safety training and programs.

What currently inhibits such an approach is the administrative cost of obtaining, maintaining and analyzing detailed records on each firm. As the social nervous system advances in information society insurance firms the administrative costs associated with such detailed records would fall greatly. Also given operational information policy insurance firms would have access to all safety features such as safety equipment, safety training and programs installed in each of their client. With variations among the states operational informational policy would also give insurance firms the right of private inspection. Such a right would privatize safety inspections.

The federal government in this research and development role would be much more effective at promoting safety than the current OSHA. For example, to reduce nerve damage from continuous, high speed input at computer keyboards OSHA is likely to be more effective promoting research and development to shift to voice recognition input than trying to define standards for existing manual keyboard input. Once voice recognition input becomes more efficient than typing input taking into consideration reduced insurance premiums, market incentives will shift technologies eliminating the old nerve damage problem. Also, transferring operational control to the states will mean greater policy variation leading to a faster rate of discovery. Variation is needed to discover the value of creating better incentives for workers to promote safety themselves as well as better incentives for employers.