CONNECTICUT GENERAL ASSEMBLY

LEGISLATIVE PROGRAM REVIEW AND INVESTIGATIONS COMMITTEE

The Legislative Program Review and Investigations Committee is a joint, bipartisan, statutory committee of the Connecticut General Assembly. It was established in 1972 to evaluate the efficiency, effectiveness, and statutory compliance of selected state agencies and programs, recommending remedies where needed. In 1975, the General Assembly expanded the committee's function to include investigations, and during the 1977 session added responsibility for "sunset" (automatic program termination) performance reviews. The committee was given authority to raise and report bills in 1985.

The program review committee is composed of 12 members. The president pro tempore of the Senate, the Senate minority leader, the speaker of the house, and the House minority leader each appoint three members.

Committee Members

Committee Staff

Carrie Vibert, Director

Catherine M. Conlin, Chief Analyst

Jill Jensen, Chief Analyst

Brian R. Beisel, Principal Analyst

Michelle Castillo, Principal Analyst

Maryellen Duffy, Principal Analyst

Miriam P. Kluger, Principal Analyst

Anne E. McAloon, Principal Analyst

Scott M. Simoneau, Principal Analyst

Michelle Riordan, Legislative Analyst II

Janelle Sagness, Legislative Analyst II

Bonnine T. Labbadia, Executive Secretary

Project Staff

Carrie Vibert

Gregory Williams, Graduate Fellow

Legislative Program Review

& Investigations Committee

Sunset Law in Connecticut

December 2007

Table of Contents

SUNSET LAW IN CONNECTICUT

EXECUTIVE SUMMARY

INTRODUCTION 1

Report Format 4

I. SUNSET REVIEW PROCESS AND CRITERIA 7

Overview 7

Types of Entities 7

Review Criteria 9

Review Process of PRI Committee 10

II. CONNECTICUT EXPERIENCE: 1979-1984 AND 1984 - PRESENT 13

Activities (1979-1984) 13

Results (1979-1984) 15

1984 to Present 18

III. SUNSET IN OTHER STATES 25

History 25

Key Aspects of Sunset Laws 27

IV. SUNSET IN CONNECTICUT: 30 YEARS LATER 29

Legislative Oversight and Where It Happens 29

Sunset Reviews and Program Reviews: A Comparison 31

Other Information on Program Review Process 35

Legislative Oversight and Performance Measurement 36

APPENDICES

Executive Summary

SUNSET LAW IN CONNECTICUT

Connecticut's sunset law was enacted in 1977 as part of a major reorganization of state government. Intended to be a cyclical, periodic review of certain statutorily specified state entities and programs with at least the prospect, if not threat, of automatic termination, the Connecticut legislature went through one five-year sunset cycle from 1979 through 1984. Since then and through January 2007, the legislature had postponed reactivation of the law five times.

Under the state's sunset law, certain statutorily specified state entities (not all state agencies) are automatically set to terminate on specific dates, unless the legislature takes affirmative action to authorize their continuation. Six months before a scheduled termination, the Legislative Program Review and Investigations Committee (PRI) is required to submit a report to the General Assembly and pertinent committees of cognizance based on a performance audit of each entity using criteria set out in the sunset law. The PRI report is required to make recommendations about abolishing, reestablishing, modifying, or consolidating each entity under review. The sunset law places the burden of demonstrating public need on the entity or program subject to termination, as well as the burden of showing that the entity serves the public interest, “not merely the interests of the persons regulated.”¹

Under the sunset law, the Government Administration and Elections Committee is responsible for holding public hearings prior to any sunset action by the General Assembly. Finally, as noted already, in order to continue an entity or program, the General Assembly must vote to do so. No action by the entire legislature would result in termination.

In January 2007, after a series of five postponements spanning almost 25 years, Connecticut's sunset law was poised to start up again with year one of a five-year cycle, in which 28 specific state entities or programs were scheduled to automatically terminate on July 1, 2008, unless they were each re-established through legislation. Uncertain of continued legislative interest in the sunset law as currently formulated, and concerned about sunset activities draining committee member and staff resources from other work absent legislative interest, the PRI committee raised legislation in 2007 to postpone the start of sunset for two years. Along with a two-year postponement, the bill, which passed, also required the committee to conduct a study on the continued need for sunset and report its findings and recommendations to the General Assembly by January 15, 2008. A central question to the discussion of resuming implementation of the sunset law was whether the landscape of legislative oversight had changed in Connecticut in the 30 years in a way that would alter views on the value of sunset as an oversight tool.

During the committee study, a number of items were reviewed: 1) the statutory sunset review process and criteria, as well as the process the program review committee followed in carrying out its mandate from 1979 to 1984; 2) Connecticut's actual sunset experience during the five years the law was active in terms of activities and results, and the sunset-related actions that have occurred since then ( primarily postponements of the law); 3) sunset laws in other states; and 4) other legislative oversight sources in Connecticut, and the differences and similarities between sunset reviews and “regular” program reviews conducted by the program review committee.

In summary, the committee found:

RECOMMENDATIONS

Introduction

Sunset in Connecticut

Under the sunset concept, statutorily specified state entities are required to periodically justify their continued existence. These entities are terminated automatically, unless the legislature takes affirmative actions to authorize their continuation for another defined period of time, or indefinitely.

In January 2007, after a series of five postponements spanning almost 25 years, Connecticut's sunset law was poised to start up again with 28 specific entities or programs scheduled to automatically terminate on July 1, 2008, unless they were each re-established through legislation.

Under the state's sunset law, the Legislative Program Review and Investigations Committee (PRI) is charged with conducting a performance audit of each listed entity or program in the year prior to its scheduled termination, over a five-year cycle. Uncertain of continued legislative interest in the sunset process as currently formulated, and concerned about sunset activities draining committee resources from other work absent legislative interest, the committee raised legislation in 2007 to postpone the start of sunset for two years. That bill passed (see Appendix A for current sunset law). The postponement was intended to allow the committee to consider the continued need for sunset and report its findings and recommendations on that need by January 15, 2008. A central question to the discussion of resuming implementation of the sunset law is whether the landscape of legislative oversight has changed in Connecticut in the 30 years in a way that alters views on the value of sunset as an oversight tool.

Connecticut's sunset law was enacted in 1977 as part of a major reorganization of state government. Under the law as originally enacted, almost 100 selected state entities or programs were scheduled to automatically terminate, staggered over a five-year period, unless each was re-created by an act of the General Assembly. The legislature specifically included a legislative finding in the law, which remains today:

In the original sunset enactment, the terminations were scheduled to begin on July 1, 1980, and continue through July 1, 1984. Figure 1 sets out the 78 entities and programs currently under the sunset law. This collection of entities and programs has changed somewhat since 1977, but not significantly. The largest category of entities regulates a profession or occupation.

Since the beginning of sunset in Connecticut, the program review committee has been charged with conducting performance audits of all the listed entities or programs prior to their respective scheduled terminations dates. As the committee had been established just five years prior to sunset's enactment to conduct program reviews in 1972, it was deemed a natural body to carry out this similar evaluation work. Three additional staff were added to committee's workforce to handle the new responsibility, bringing the total number of staff up to 13.

The committee completed its initial round of evaluations in December 1979 and, as required by the law, submitted its findings and recommendations to the Government Administration and Elections Committee (GAE) by January 1, 1980 for its consideration during the 1980 session of the General Assembly. A full cycle through the list was completed during the 1984 legislative session. During the first five-year cycle, most of the entities or programs that were reestablished by the legislature were also placed back on the sunset list, meaning they were re-created for another limited period of time (i.e., five years). Since then, due to five postponements described in more detail in Chapter II, no further sunset review cycles have been carried out.

In 1998, the PRI committee authorized a study very similar to this, for similar reasons. That study, which presented a comprehensive summary of what happened during the sunset review days of the early eighties and recommended continuing sunset in a modified form, is utilized extensively in this review. Two noteworthy events happened in 1999, the year after that sunset study was completed.

First, as a result of the 1998 sunset study, the program review committee raised, reported out, and referred a bill to GAE to implement its recommendation to modify sunset to focus exclusively on regulatory entities. That bill died at GAE during the 1999 legislative session.

Second, also in 1999, the program review committee began a study of performance measurement in Connecticut.² One of that study's results, adopted by the program review committee in December 1999, was to recommend the repeal of sunset in order to use those resources to assist in carrying out a new comprehensive approach to performance measurement.

Performance measurement in government was a subject that had caught on a few years earlier in Connecticut and other states and, in addition to being seen as a benefit to better executive branch management, was also seen as leading to more effective legislative oversight. In its performance measurement study, the PRI committee concluded that to kick start performance measurement in Connecticut, which existed on paper only, the program review committee should be involved on behalf of the legislature in analyzing and commenting on agency performance measure and benchmark data prepared by the agencies under the Office of Policy and Management's guidance. To ensure the program review committee could continue doing its normal studies as well as this new task, the committee recommended that the sunset law be repealed, noting “this law consumes an enormous amount of staff and legislative time and, as currently written, is disproportionately focused on relatively small and narrow programs.”³

The bill to implement the performance measurement study recommendations, including the sunset repeal, was raised in the 2000 legislative session, reported out by three committees, but amended in the Senate to a much scaled back version. The amended version required OPM to report on how it would establish a performance measurement program for state agencies, instead of actually implementing such a program. The amended version passed the Senate but died in the House.

These two events -- an unsuccessful proposal to continue but modify sunset, followed by a proposal, also unsuccessful, to promote the actual implementation of government-wide performance measurement with the assistance of the program review committee, and repeal sunset -- took place within a year of each other. While they are conflicting proposals, they are similar in suggesting a search for a new direction in legislative oversight that does not include the current sunset law. While the lack of success of those proposals might indicate an underlying interest in the current sunset law, the fact that those two events were followed by further postponements casts doubt on that interpretation. After reviewing the nature and scope of the program review committee's work over the last several years in comparison to the actual experience of sunset, and recognizing possible new avenues to enhance legislators' opportunities to know how well the programs they established are actually working, the program review committee concludes that the landscape of legislative oversight has changed in Connecticut since the advent of sunset, and recommends Connecticut's sunset law be repealed.

Methodology

The history of the sunset law nationally and Connecticut's own experience were reviewed in this study, including an examination of the actual Connecticut sunset reviews conducted in the late 1970s to the mid-1980s and PRI committee meeting minutes from that time period. A 1999 PRI study of the sunset law and process, entitled Sunset Review Process in Connecticut, was used as a reference, as was a 2000 PRI study entitled Performance Measurement. Selected states with continuing sunset activities were researched.

Report Format

This report has four chapters in addition to this introduction. Chapter I describes the statutory sunset review process and criteria, as well as the process the program review committee followed in carrying out its mandate from 1979 to 1984. Chapter II examines Connecticut's actual sunset experience during the five years the law was active, and sets out the sunset-related actions that have occurred since then, primarily postponements of the law. A brief overview of sunset laws in other states is provided in Chapter III. Finally, Chapter IV identifies other legislative oversight sources in Connecticut, compares the differences and similarities between sunset reviews and “regular” program reviews conducted by the program review committee, and sets out the committee recommendations.

Agency Response

It is the policy of the Legislative Program Review and Investigations Committee to provide agencies subject to a study with an opportunity to review and comment on the recommendations prior to publication of the final report, for inclusion in the report. As the subject matter of this study potentially affects so many agencies, this was not practicable.

Chapter I

Sunset Review Process and Criteria

Overview

Public Act (P.A.) 77-614 defined the process for the General Assembly to follow to carry out its sunset responsibilities (codified at C.G.S. Sec. 2c-1 et. seq). The process, depicted in Figure I-1, begins with the program review committee conducting a review of the entities or programs scheduled to terminate under C.G.S. Sec. 2c-2. At a minimum, the review must address the criteria outlined in C.G.S. Secs. 2b-7 and 2b-8. At the conclusion of its work, the program review committee is required to produce a written report for each entity or program reviewed that includes a recommendation to terminate, reestablish, or modify the entity or program.

During the second phase of the sunset review process, the Government Administration and Elections Committee and the committees of cognizance over the entities or programs under review receive the reports prepared by the program review committee. The recommendations of the program review committee serve as the focal point of the public hearing GAE is required to hold on each of the entities or programs scheduled to terminate. Historically, the committees of cognizance also raised legislation reflecting the recommendations of the program review committee, held public hearings, and proposed their own recommendations in the form of committee bills.

The Government Administration and Elections Committee is charged by statute with making the official sunset recommendation to the General Assembly (C.G.S. Sec. 2c-6). In making its decision, GAE is not bound by the recommendation of the program review committee or any of the committees of cognizance. GAE's recommendation to reestablish or modify an entity or program is sent to the full General Assembly in the form of a bill. If the intent of GAE is to terminate an entity or program, no bill is required.

The last phase in determining the fate of an entity or program subject to a sunset review is governed by the requirements of the normal legislative process. As Figure I-1 shows, an entity or program can only be reestablished if the General Assembly passes a bill, and it is signed by the governor, or the legislature overrides a veto issued by the governor. Further, if the legislature wants to put the reestablished entity on the sunset list again, subject to another termination date within a five-year period, affirmative legislative action is also needed.

Finally, if the General Assembly fails to pass a reestablishment bill or override a veto, the affected entity or program is given a one-year wind down period to conclude its affairs before termination.

Types of Entities

The original law emphasized small regulatory boards and commissions, but did include some large agencies and programs. A total of 94 entities and programs were covered. Virtually everything with the word “board” or “commission” in its title was selected for inclusion, as were

a few other entities and programs that had, for one reason or another, caught the attention of the proponents of reorganization.

The entities and programs included under the original sunset mandate can be divided into five broad functional categories, as can the entities and programs on the current list. The categories are based on the general purpose of the entity or program and include:

Review Criteria

Similar to sunset laws in other states, Connecticut's law lists criteria that are to be considered in reviewing the merits of an entity or program, although the law does not limit the program review committee to the criteria. As outlined in C.G.S. Sec. 2c-7, the criteria for determining whether there is a public need for the continued existence of an entity or program shall include among other things:

In addition, C.G.S. Sec. 2c-8 requires the following criteria be considered in determining whether the general public and not just the persons regulated have been served by any entity or program that exercises regulatory authority:

PRI Committee Review Process

As noted above, the program review committee is responsible for conducting reviews of each entity or program covered by the state's sunset law. The committee is required to complete the review by January 1 of the year the entity or program is scheduled to terminate. Figure I-2 outlines the sequence of actions the committee undertook in meeting this responsibility during the one completed round of sunset reviews from 1979-1984.

Chapter II

Connecticut Experience: 1979-1984 and 1984-Present

Activities

Figure II-1 shows the number of entities or programs reviewed under Connecticut's sunset law from 1979 until it was suspended after the completion of the fifth and final round of the initial cycle. The variation in the number of sunset reviews the committee performed per year is noteworthy and significantly affects the committee's operation.

Figure II-2 shows the distribution of the 94 sunset reviews among the five functional categories described in Chapter I. Although regulatory related entities and programs comprise a majority of those on the sunset list -- 52 percent when "regulation of individual practitioners" and "enforcement of state standards" are combined -- coverage under Connecticut's sunset law is not limited to regulatory matters. Indeed, nearly a quarter of the reviews (23 percent) involved entities classified as providing a service directly to the public or to other state agencies.

Resources Expended

Table II-1 tracks the effort expended by the PRI committee and its staff over the life of the sunset cycle. The table includes annual data on the number of reviews completed; staff assigned to sunset; committee meetings held; reviews per assigned staff person; and committee meetings per review.

The data in the table show output per assigned staff increased and committee meetings per review declined over the five-year cycle. There are two explanations for this increase in efficiency. First, the reservoir of knowledge accumulated as the committee and staff progressed through the sunset review cycle resulted in less time being needed to: develop methods; understand the basic environment within which all of the entities and programs had to operate; identify problems; and conceptualize solutions. This resulted in the need for less staff time and fewer committee meetings to grasp and resolve issues surrounding the reviews.

Second, the final two years of the sunset cycle had proportionally more entities and programs falling into the "advisory" and "policy guidance" categories than the earlier years. Such entities and programs are typically less complex in their structure and operation than those enforcing state approved standards or regulating licensed practitioners and, therefore, require less time and effort to review.

The resource information in the table is limited and the absence of certain key resource data from the table should be noted. Specifically, the table does not contain information on the time spent on sunset-related activities by program review committee staff not directly assigned to sunset (i.e., staff director, staff attorney, and clerical staff). Also missing from the table is the time spent on sunset activities by other legislative staff, committees, and the full General Assembly. The limited availability of records and difficulties encountered in reconstructing data from the records that could be located made it impossible to present other staff activity data covering the entire sunset cycle.

However, a one-time snapshot is available with respect to resources expended by the program review committee staff. At the completion of the first year of the cycle (1979), an estimate of the committee staff time spent planning, conducting, organizing meetings, presenting materials, and writing reports was developed. The estimate covered analysts conducting the reviews plus all other program review committee professional and support staff involved during the first year. The estimate put staff time at 1,039 total person-days or approximately 50 person-days per review.

Results

During the completed five-year sunset cycle, the program review committee examined 94 entities. In total, the committee proposed slightly more than 350 recommendations, of which approximately 270 were implemented. The recommendations ranged from the obscure --elimination of the regulation of second-hand hats -- to the significant -- restructuring the state's hospital regulatory program. The key results of Connecticut's sunset review process include:

Given the initial focus of the sunset review process on the termination of entities and programs, it is appropriate to look at Connecticut's experience in this area. Overall, the program review committee proposed terminating 32 entities. Seventeen of these recommendations were acted upon favorably by the General Assembly; the other 15 were rejected.

Figure II-3 illustrates the relationship between the committee's recommendations to continue or terminate entities within each of the five functional categories previously identified. The graph shows the highest number of termination proposals occurred in the “regulation of practitioners” classification (14). In terms of the percentage of reviews resulting in termination recommendations, two categories stand out: “policy guidance” (55 percent) and “advisory” (50 percent). The high number of terminations proposed by the committee in the “regulation of practitioners” category is directly related to the statutory review criteria. In all of these instances the committee, adhering to the criteria, concluded such things as a bad haircut, dead tree, or poor landscape job did not endanger the public health, safety, or welfare. In the committee's view, regulation of practitioners, in areas where a reasonable consumer had the capability to assess and assume the risks, needlessly restricted competition.

The high percentage of committee-proposed terminations in the policy guidance and advisory categories also is not surprising when the statutory review criteria are considered. It is very difficult for an entity or program designed to provide policy guidance or advice to prove its elimination would endanger the public health, safety, or welfare. Thus, despite low costs, their existence could not be justified, and the committee was compelled to recommend termination. It should be noted such recommendations were typically met with outcries of objection by interests represented by the involved entities.

As noted above, the committee proposed 32 terminations, but the General Assembly adopted only 17. Figure II-4 illustrates by functional category the relationship between the termination recommendations of the program review committee and subsequent actions of the General Assembly. The graph shows the General Assembly concurred with slightly less than 30 percent of the committee's termination recommendations in the “regulation of practitioners” category, about 60 percent in the “policy guidance” and “advisory body” categories, and all five recommendations in the “enforcement of state standards” and “service” categories.

Although not shown in Figure II-4, the General Assembly concurred 100 percent of the time when the program review committee recommended an entity or program be continued. A closer examination of the 17 committee termination proposals agreed with by the full General Assembly reveals these decisions were not as significant as might first be thought. Six of the 17 agreeing actions -- including three of the four agreeing actions in the enforcement of state standards classification -- involved consolidation of boards and commissions. In five instances, the authority for the regulatory or service function underlying the terminated entity was transferred to another state agency. The activity level of three of the remaining six entities terminated was virtually nonexistent.

One explanation for the differences between the committee and the General Assembly can be found in the decision-making processes each followed. The committee was guided by the statutory criteria and had sufficient staff and time to obtain and analyze data related to claims made by the involved entities and their supporters. This enabled committee members to reach an acceptable comfort level with the rationale for terminating an entity or program.

On the other hand, the full General Assembly relied heavily on obtaining information from public hearing testimony and direct contact with constituents. It had little time to sort facts from claims. As a result, its members were very aware that terminating an entity often meant imposing a real or perceived negative consequence (e.g., loss of ability to control competition, loss of influence, etc.) on a portion of the state's population. Legislators also seemed to sense many terminations would only provide an incremental benefit that would be little noticed or appreciated by most citizens.

As the key outcomes listed at the beginning of this chapter indicate, there was more to the sunset process in Connecticut than terminating entities and programs. Indeed, nearly three out of every four sunset-related recommendations made by the program review committee involved issues other than the continuation or termination of an entity or program. Included were recommendations to limit matters subject to regulation, change the level of regulation, streamline and standardize procedures, fine tune the 1977 reorganization by transferring functions from one state agency to another, and enact a host of reporting requirements designed to increase accountability by making information available to the legislature and public.

Figure II-5 shows the number of such proposals introduced by the committee in each functional area. Also included in Figure II-5 is the number of proposals adopted by the General Assembly. The overall number of recommendations per entity reviewed ranged from a low of about 1.3 in the “policy guidance” classification to a high of 3.5 in the “service” category.

Most notable, the agreement rate between the committee and the full General Assembly was much higher in areas where the issue did not involve whether to terminate an entity or program. It ranged from 59 percent in the “enforcement of standards” category to 92 percent in the “policy guidance” area. Overall, the rate of agreement on recommendations not addressing termination issues was nearly 75 percent.

In summary, if the success of Connecticut's sunset program is measured by the number of entities or programs terminated, the results are mixed. Some progress was made in eliminating entities on the sunset list, but it fell short of what was envisioned by proponents of a sunset law. However, if success is measured by the number of recommendations resulting in laws, the experience has been decidedly more positive. Further, if the significance of the changes adopted as a result of the sunset process is considered, the effort was very successful.

Two examples of such changes involve the regulation of business practices for health professionals and the requirements to be a manicurist. In the case of the former, regulation of business practices of health professionals (e.g., number of offices, advertising, etc.) was separated from matters of competence and removed altogether from the control of practitioners of the profession. With respect to manicurists, requirements they receive 500 hours of training and be restricted to practicing only under the supervision of a licensed cosmetologist were eliminated. This gave rise to the large number of nail boutiques currently located throughout the state.

1984 to Present

Although the sunset law has not been operational since 1984 due to its many postponements, sunset-related activity has occurred since then. This activity falls into two categories: changes to the sunset entity list and sunset law postponements.

Sunset list changes. In regard to changes to the list, a handful of entities has been added to the sunset list since the first postponement in 1983. None have been reviewed yet due to the postponements. These are in addition to a few entities added to the review list prior to the first postponement but, due to the timing of their additions, have also never been reviewed. Finally, a few entities that were reviewed during the original sunset cycle and re-established were not placed back on the sunset list to be reviewed in another five years (e.g., the Connecticut Student Loan Foundation and the Connecticut Agricultural Experiment Station).

The sunset law has been postponed six times, including most recently in 2007 (see Appendix B). Selected aspects of the postponements are discussed below.

First postponement (1983). The first postponement of sunset was enacted during the 1983 legislative session, as the committee was winding up the first five-year cycle. (In 1983, the committee was set to conduct performance audits of 22 entities scheduled to terminate on July 1, 1984.) In 1984, the sunset cycle was to begin again, with PRI required to review 19 entities, most of which involved regulated health-related providers previously assessed in 1979. As reflected in PRI meeting minutes, the committee generally agreed that instead of repeating the list, a new schedule of entities should be proposed. At the same time, Senator Joseph Harper, who was one of the Appropriations Committee's co-chairs that session and a recent past PRI co-chair, had raised a bill to shift the focus of sunset from regulatory boards and agencies to major operating departments for the next five years, targeting five to 10 major agencies for performance audits over those years. The program review committee endorsed the Harper bill.

What ultimately passed (P.A. 83-466) instituted the first five-year postponement of sunset and a one-year pilot to test the Harper approach. Specifically, the existing sunset list, which had entities terminating each year out to July 1, 1988, was repealed. A new list, almost identical to the repealed list, was put in place, with a beginning termination date of July 1, 1990.

Also, a one-year pilot to test the idea of having the program review committee examine large agencies as opposed to predominantly regulatory boards was also established, with the following provisions:

As a result of the pilot, three separate PRI reports were produced in 1984 on three DIM programs: 1) Error Detection and Prevention; 2) General Assistance Program; and 3) Agency Management. While none of the programs were defunded, several recommendations were made. The idea of changing the sunset focus permanently to larger agencies was not pursued.

Second and third postponements (1988 and 1993). In 1988, with committee work on sunset slated to begin in 1989 again for the first round of entities (with July 1, 1990 termination dates), P.A. 88-165 set back the sunset law for five years with a new beginning termination date of July 1, 1995.⁴

The same thing happened in 1993 via P.A. 93-250, which pegged the restart of the sunset law back another five years to July 1, 2000. This bill addressed other committee matters, including clarifying its authority to receive and review confidential state records. In regard to the sunset delay, then-PRI co-chair Representative Jonathan Pelto testified before the GAE committee that the activation of the sunset review process was “a policy matter as much as it is anything else” for GAE and the legislature, but that in order to conduct the sunset reviews and continue the other program review work of the committee, additional staff would be needed. While he recommended a 10-year postponement, another five-year delay was instituted.

Fourth postponement (1998). In 1998, the program review committee was again faced with the restart of sunset, and in addition to seeking a delay, it determined that a study should be done on the need for and benefits of sunset and alternative ways of addressing the needs. Unlike the previous delays, P.A. 98-30 pushed off the start for just three years, and a study was conducted in 1998.

The report, completed in December 1998, reviewed Connecticut's experience with sunset as well as those of other states. The report noted the reasons frequently cited for repealing or suspending sunset laws in other states, which included:

Based on its Connecticut-specific findings provided in Table II-2, as noted earlier in this chapter, the program review committee concluded if the success of Connecticut's sunset program was measured by the number of entities or programs terminated, the results are mixed. Although progress toward the goal of sunset was made, it fell short of what was envisioned by proponents of the sunset law. However, if success was measured by the number of adopted recommendations resulting in new laws leading to improved operations of entities and programs, the experience has been decidedly more positive. Ultimately, the committee concluded that the sunset concept was a valuable oversight tool and should remain available for use by the General Assembly.

It could be argued that many of the benefits attributed to the sunset process in the 1998 committee findings that supported continuation could also have occurred through program reviews. Perhaps implicit in the committee findings is a belief that the sunset successes are distinguishable from those potentially achieved by program reviews. This belief is in part because of sunset's action-enforcing mechanism (i.e., an automatic termination date), which requires asking whether the entity is needed at all, and forces the legislature to take affirmative action if the answer is yes.

In assessing the form in which sunset should continue in 1998, the committee considered four options. Significantly, no option maintained the status quo. Three options maintained the automatic termination tool that makes sunset reviews unique; the fourth option did not. The latter essentially sketched out an alternative PRI function more closely allied with the budget. Under all options, the review cycle was lengthened.

The option adopted by the committee in 1998 was to continue sunset but with modifications, as set out in Figure II-6. During the 1999 legislative session, PRI raised the recommendation as a bill (H.B. 1177), which died in GAE.

At first glance, the proposal might seem to limit the scope of sunset by eliminating three out of five types of entities from the list. All entities and programs whose primary purpose was to provide advice, policy guidance, or direct services would be removed from the list. Left were entities regulating individual practitioners or enforcing state-approved standards, to which would be added similar entities not already on the list. The proposal as it was drafted into legislation focused on the processes of regulation rather than entities, and included a broad definition of regulatory process, called “state action”, which meant:

This “state action” definition in the proposal in fact greatly expanded the scope of sunset.

While the proposal gave the program review committee discretion in regard to which entities carrying out “state actions” would be reviewed in detail, it effectively created a mandatory reauthorization requirement for the Connecticut legislature for many state programs. Whether or not the program review committee examined any process on the list, the authority for the process (i.e., state action) was scheduled for termination on a date certain unless it was affirmatively reenacted by the legislature.

On another front, in 1999, just one year after the committee reviewed the sunset law, the committee directed its staff to conduct a study of performance measurement in Connecticut. The result of that study was to recommend a new and revitalized system of performance measurement, building on the many parts already on the books in Connecticut but not implemented by the executive branch. The program review committee was to be involved in the new system. The legislation containing the study recommendations was not successful, due in part to opposition from OPM. The performance measurement study is pointed out here because the committee in that report recommended that the sunset law be repealed:

The ninth recommendation calls for the repeal of the state's sunset review law. The committee believes if the sunset law is not repealed, the demands it places on the program review committee coupled with the requirements of the above recommendations will force the committee to either severely reduce the number of studies it undertakes annually or add staff. In the opinion of the committee, the expense of adding staff could be avoided and the legislature better served by repealing the sunset law. This law consumes an enormous amount of staff and legislative time and, as currently written, is disproportionately focused on relatively small and narrow programs.⁵

Fifth postponement (2001). Without an alternative approach in place after the 1998 study, the sunset law was again set to restart with a termination date of July 1, 2003, for the first round of entities. In 2001, P.A. 01-160 was passed with another five-year postponement, setting back sunset to July 1, 2008 (as the date of the first terminations).

Sixth and current postponement (2007). Early in the 2007 legislative session, the newly formed 2007-2008 program review committee was confronted with the sunset question again. With a first-year termination date of July 1, 2008, the committee would have to conduct 28 sunset reviews during calendar 2007 in order to have the required reports prepared for the General Assembly and the GAE committee by January 1, 2008, pursuant to the sunset law. The numerous postponements over the years with little comment suggested a low level of interest in the sunset process, but because there also seemed to be a reluctance to outright repeal the law, the committee determined to try again to clarify the legislative interest in sunset. Thus, P.A. 07-33 set back the start date for termination to July 1, 2010, the shortest postponement ever enacted. Conducting the current study was a required part of the delay.

Figure II-6. PRI Recommendation from 1998 Sunset Review Study

Connecticut's current Sunset Law should be modified in the following manner:

Chapter III

Sunset in Other States

Over the years, as many as 38 states have enacted sunset laws. Today, 19 states have active sunset programs of varying scope.

History

When Connecticut adopted its sunset law in 1977, it was among the leaders of what has been called the “Sunset movement.”⁶ Between 1976 and 1982, the peak of sunset, 36 states enacted some version of a sunset review process. The advocacy group Common Cause was at the forefront of promoting state sunset laws as a way to increase legislative oversight, specifically “to provide the incentive and discipline necessary to motivate public officials to increase program evaluation”. ⁷

A 1986 report by the South Carolina State Reorganization Committee surveyed states' experiences with sunset, and offers an instructive overview of the multiple intents behind sunset, explaining that “[s]unset was the product of at least three streams of legislative reform activity of the 1970s:

The report goes on: “It was thus under several banners that sunset was launched, a favorite of many diverse legislative interests, and from the outset something of a victim of exaggerated expectations and misunderstandings.”⁹

Over time, 13 of the original 36 states repealed their sunset statutes, and six states (including Connecticut) suspended their sunset laws in some fashion, leaving 19 states with statutory sunset processes that are actually active. Twelve states have never enacted sunset laws.

Among the reasons frequently cited for repealing or suspending sunset laws are:

The South Carolina report contained similar conclusions:

Current status. Despite this sunset retrenchment, as noted above, 19 states still have active sunset laws, but of varied scope. Key aspects of these active sunset laws are highlighted below. First, a brief description of the Texas sunset law is provided, as it is perhaps the most well known sunset process currently.

Texas, which established a sunset law the same year as Connecticut in 1977, still has its sunset law as the backbone of legislative program evaluation in Texas, in addition to the state's budget process. Texas takes a comprehensive approach to sunset -- about 130 state agencies are subject to the Texas Sunset Act, covering almost all the large and small agencies of Texas state government. Each of these agencies is scheduled to terminate on a date certain unless reauthorized by the legislature. Typically, agencies go through a sunset review every 12 years, although the review schedule may be changed if the legislature wants more frequent review of a particular agency.

The Sunset Advisory Commission is the entity that oversees the sunset process, a 12-member body of 10 legislators (five Senate and five House members) and two public members. The commission is assisted by a staff of 30. The Texas statute, like Connecticut's, contains a set of criteria for the reviews.

Even though the Texas sunset law is 30 years old, it is interesting that in a published Guide to the Texas Sunset Process (January 2006), the commission notes how sunset reviews have evolved in Texas: “Early sunset reviews focused on whether the state should regulate certain occupations and how to ensure an arm's length relationship between regulators and the regulated. As the sunset process has matured over time, both the commission and the legislature have focused on substantive changes to the policies carried out by state agencies.”¹¹

Key Aspects of Sunset Laws

Sunset laws differ among the states. The presence of an automatic termination date, requiring affirmative legislative action to override, is the hallmark of a sunset review law - an “action-forcing mechanism, carrying the ultimate threat of elimination.” A related core sunset component is that prior to the termination date, there is a mandatory review of the entity or program using established criteria to conduct the review.

Indeed, a few states that are counted among those that have repealed or not enacted sunset laws have created “quasi-sunset” systems, whereby they have established the mandatory review of entities and programs scheduled at certain, pre-set times, but do not include the “action-forcing mechanism of an automatic termination date (i.e., Maine, Vermont, Virginia, and West Virginia).

The key areas of variation include selection of entities and programs to be covered by a sunset provision, review criteria, length of time between reviews, staffing, and responsibility for overseeing the conduct of the reviews and preparing recommendations. However, the literature suggests that some states that have continued with sunset have modified their processes by: lengthening the time between reviews; reviewing larger agencies; and building in flexibility in terms of what gets reviewed and at what level.¹²

Selection/Coverage. In terms of selecting the type of entities and programs covered, state sunset laws can be divided into four types. Included are those:

Further, some states focus on new programs versus existing programs.

Review criteria. While the exact wording of the criteria used to evaluate entities and programs under sunset laws varies among the states, two broad categories can be identified. The first deals with criteria aimed at assessing the need for the state to be involved in the area under review and the appropriate level of involvement. The second category concerns matters relating to performance, including the extent to which goals have been met and resources efficiently used.

Timing of reviews. Another significant difference between states is the timing of their schedule of sunset reviews. Review cycles range between as low as four years, to as high as 15 years.

Staffing. Staff for conducting sunset reviews have generally come from one of four basic sources including:

Responsibility for conducting reviews. In many states, special legislative sunset committees have been established to make the policy choices mandated under their respective sunset laws. Several states have assigned this responsibility to standing committees, especially those committees with jurisdiction over budget matters.

Existing oversight committees have also been a popular choice for this task. A few states have created special commissions composed of legislators and citizens to oversee the process and make sunset recommendations.

Table III-1 shows the breakout of states with active sunset processes by topic coverage.

Chapter IV

Sunset in Connecticut: 30 Years Later

The sunset law and process that is the subject of this study was enacted 30 years ago. Almost twenty-five years have passed since the end of the first and only cycle of completed sunset reviews, with six subsequent postponements. A central question to the discussion of continuing the sunset law is whether the landscape of legislative oversight has changed in Connecticut in those 30 years in a way that alters views on the value of using sunset as an oversight tool. Specifically, the view that matters most is that of legislators, who need to feel that sunset is a meaningful tool for them.

One of the attractions of sunset in the 1970s was that it provided a tangible mechanism to strengthen legislative oversight at a time when state legislatures had little to no resources of their own available to them to assess what happened to the laws they passed. Instead, legislatures were more dependent on the executive branch for implementation information. That balance has shifted since then as legislatures developed increased independent research and evaluation capacity, primarily through increased professional staff.

This chapter provides a brief overview of the sources of oversight in Connecticut, with emphasis on two aspects: 1) the function and experience of the program review committee in carrying out its “program review” charge, and comparing that to the sunset function and experience, which also involves the committee; and 2) the growing interest and use in Connecticut of performance, or results-based, measures as a tool for legislative oversight, particularly demonstrated by the activities of the Appropriations Committee in the last two years. The program review committee believes that comparing the scope and nature of the oversight work the program review committee does currently with Connecticut's sunset process and experience, and considering how current program review work could be augmented in the future with an orientation toward the results-based accountability approach that is starting to be used in the budget process, the usefulness of returning to sunset seems extremely limited.

Legislative Oversight and Where It Happens

The legislative oversight function is usually distinguished from enacting laws, appropriating funds, and representing constituents, although legislative oversight can occur through all those activities. For purposes of this report, the focus is on formal (vs. informal) oversight, where the purpose of an activity is to find out about and evaluate how an agency or program is operating. The definition of “program review” in the program review committee statute offers a definition of formal legislative oversight:

In Connecticut, formal oversight may be exercised by several different entities:

Deciding how much oversight is enough and in what form is a policy decision for a legislature, with multiple considerations in play. On the one hand, at least one observer believes Connecticut legislators are not interested in doing the work of oversight. In Under the Gold Dome An Insider's Look At the Connecticut Legislature, Judge Robert Satter wrote in 2004:

On the other hand, the National Council of State Legislatures, in a 2005 briefing paper on legislative oversight in the states, noted the increasing need for legislative oversight.

Legislatures face several challenges in conducting oversight, including the never-ending tug of war that characterizes the separation of powers between the legislative and executive branches. It can be difficult to obtain valid data from agencies, governors can resist oversight as “micromanagement” and demands for studies can overwhelm the capacity of legislative evaluation offices. It can also be difficult to structure processes that effectively use oversight information in the committee and budget processes. Nonetheless, given the increasing scope of policy challenges facing legislatures under federal devolution and ever-growing budget constraints, the need for effective legislative oversight will continue to grow.¹⁴

Sunset Reviews and Program Reviews: A Comparison

The sunset law has been dormant for almost 25 years, meaning the program review committee has not been conducting any sunset reviews. However, the program review committee has been active in examining state government programs and agencies to assist the legislature. This point is made because a review of the House and Senate floor debates over the years on the sunset postponement bills could leave the impression that the sunset law is the only vehicle for legislative oversight and, without sunset, no oversight is occurring in Connecticut. Underlying that impression, although not quantified by the program review committee, may be a misconception of what the Connecticut sunset law entails or what the actual experience was (which would be understandable as a generation has gone by since the last sunset study).

Thus in deliberating about what to do about sunset, it is useful to understand the differences and similarities between the sunset process and experience (governed by C.G.S. Secs. 2c-1 to 2c-12) and the “regular” program review process and experience (governed by C.G.S. Secs. 2-53d to 2-53j). The differences and similarities can be categorized as those based in statute and those based on practice. The statutory differences between sunset reviews and program reviews involve: 1) how study topics are determined; 2) the type of study topic; 3) evaluation criteria; and 4) the nature of post-study legislative activity. Table IV-1 sets out the statutory differences, and also discusses relevant committee practice. In practical effect, as the sunset experience turned out in Connecticut, there is not much substantive difference between the process of either a sunset review or a program review, or between the results of each type of review.

In regard to any differences based on post-study legislative activity provisions — particularly the sunset requirement for a mandatory vote to re-establish an entity — it is hard to say based on the actual sunset experience what ultimate difference that made, given that most entities were re-established. The only legislation the General Assembly is required to enact is the budget. A key question is how beneficial it is to the General Assembly to be required to enact legislation to continue entities that, while clearly not unimportant, are narrowly focused.

In summary:

The program review committee recommends that the sunset law be repealed.