Desktop version

Home arrow Law arrow Constitutional Sunsets and Experimental Legislation: A Comparative Perspective


The burdens of the implementation of sunset clauses and experimental legislation designed to produce information and test the effectiveness of new laws and regulations can be compared, to a certain extent, to those of conducting empirical studies: both are costly and time-consuming, remaining fairly unused by lawyers.[1] The enactment of sunset clauses and experimental legislation is burdensome in multiple ways.

Guido Calabresi, in his book A Common Law for the Age of Statutes, evidenced his scepticism towards sunset clauses, arguing that the implementation of this instrument would require significant resources.[2] In fact, the literature has been highly critical as to the employment of sunsets in determined fields, such as tax law, since these legislative instruments would misrepresent the true costs of legislation.[3] It has been argued that sunset clauses employed in the context of production tax credits designed to stimulate investment in renewable energy can ‘frustrate the simplicity goal of the tax system by increasing the costs of compliance’.[4] If the sunset is frequent (as in the mentioned case), taxpayers must spend additional resources in order to ensure that their projects are completed before the sunset date in order to obtain the production tax credit and avoid being submitted to new rules which differ from the previous ones.

According to the study performed by Kearney in 1990, most states in the United States did not have any real perception of the actual costs of sunset clauses. Whereas sunset clauses should theoretically lead to savings since they imply the termination of agencies; in practice, Kearney concluded that it was difficult to calculate if there were effective savings and if they compensated for the costs implied by sunset clauses.[5] The excessive costs of sunset review, often outweighing the benefits, have been difficult to calculate since the enactment of the first sunset clauses. No substantial improvements at this level have been described more recently.[6] However, this ‘story’ can have a different ending, provided that we analyze the costs of the lawmaking process as a whole and enact sunset clauses within a clear framework. Temporary legislation does not necessarily imply higher costs: such instruments simply ‘allocate the costs between the current-period and future-period legislature’. Jacob Gersen affirms that ‘the higher transaction costs of temporary legislation during the sunset year may be partially, if not entirely, offset by lower relative transaction costs in the time periods prior to the sunset’. According to Gersen, the intuition that temporary legislation is ‘more costly’ is not supported by either theoretical or empirical claims.[7]

In Germany, it was also claimed that sunset clauses also appeared to be a costly instrument since they lead to savings in the first years of implementation, due to the reduction in the number of regulations; but later, as more legislative oversight was required, the costs would increase.[8] It is worth remembering that a number of German states have adopted general sunset clauses, meaning that a significant number of regulations must be extinguished. In order to avoid the unwise extinction of regulations, states will have to incur ‘legislative oversight’ costs.

There is scarce literature on implementation costs of experimental legislation; however, this does not mean that these legislative instruments have not received the same objections as sunset clauses. The implementation of experimental laws implies transaction costs of informing judges and lawyers about the temporary experimental rules and training them in the introduced specificities; (delay) costs of the application of beneficial rules to those that are not included in the sample group; the implicit costs of potential errors of implementing ineffective rules; and monitoring and evaluation costs.[9]

Although temporary laws require ‘multiple rounds of enactment costs’ and permanent laws do not, this does not mean that the former will be more costly than the latter, since all laws must fulfil the same constitutional, legal and procedural enactment requirements and respective costs.[7] It is difficult to conclude whether sunset clauses and experimental laws are more costly than permanent laws and particularly inefficient ones: the maintenance costs of temporary laws may be lower because they last, in principle, fewer years and no repealing costs will be incurred when the sunset is approaching.[11] However, the efficient monitoring of sunset clauses and experiments can undoubtedly counterbalance the mentioned lower costs, being determinant in a cost-benefit analysis of these instruments.

In conclusion, sunset clauses and experimental legislation require a clear enactment and implementation framework. Notwithstanding the discussed objections, legislators should not be discouraged by the examples of ineffective laws that were passed and implemented without explicit goals, for an insignificant duration or without an adequate evaluation methodology. Sunset clauses and experimental laws can perform differently and better.

  • [1] Laurens Walker, ‘Perfecting Federal Civil Rules: A Proposal forRestricted Field Experiments’ (1988) 51 Law and Contemporary Problems 67,70.
  • [2] Guido Calabresi, A Common Law for the Age of Statutes (CambridgeUniversity Press, 1992) 61.
  • [3] William G. Gale and Peter R. Orzag, ‘Sunsets in the Tax Code’ (2003) 99Tax Notes 1153, 1157.
  • [4] Dewey, ‘Sundown and You Better Take Care’, n. 111 above, 1122.
  • [5] Kearney, ‘Sunset: A Survey and Analysis of the State Experience’, n. 90above, 54.
  • [6] Davis, ‘Review Procedures and Public Accountability in Sunset Legislation’, n. 105 above, 396.
  • [7] Gersen, ‘Temporary Legislation’, n. 51 above, 263.
  • [8] Bertelsmann Stiftung, Sunset Legislation and Better Regulation, n. 59above, 19.
  • [9] Walker, ‘Perfecting Federal Civil Rules’, n. 137 above.
  • [10] Gersen, ‘Temporary Legislation’, n. 51 above, 263.
  • [11] Ibid. 247.
< Prev   CONTENTS   Source   Next >

Related topics