SJR 350

Commission on the Commonwealth's Planning and Budget Process

December 19, 1997, Richmond

The final 1997 meeting of the Commission on the Commonwealth's Planning and Budgeting Process focused on North Carolina's legislative general fund financial model. The computer model is used during the legislative budget review process to project impacts on general fund expenditures and revenues over a 10-year period. By helping to establish a frame of reference, the model provides a view of alternative budget proposals.

North Carolina's Legislative Model

The model was developed in 1992 for the North Carolina General Assembly by the Barents Group, an affiliate of KPMG Peat Marwick. Run on a personal computer, the model is based on the concept that the level of services will continue at the level of the current budget year, and expenditures associated with those service levels are adjusted for inflation and other drivers. The base year is updated annually to provide a rolling 10-year projection.

A senior fiscal analyst with North Carolina's Fiscal Research Division stressed that the model is a simulator, not a predictor, of revenue and expenditure changes. Given the future economic and demographic outlook, the model provides an effective way of simulating changes if the budget is allowed to grow in the future as it has in the past. The value of the long-range projections is the ability to spot trends, and the model is not intended to predict future revenue and expenditures with precision. As such, it is analogous to a performance statement used in private businesses.

A characteristic of the model is its incorporation of expenditure drivers for major categories. For example, the projections in the human services category incorporate projected changes in unemployment and demographics. The model incorporates, rather than duplicates, expenditure projections developed by executive branch agencies.

Using stated assumptions about the growth rates in such elements as state population, school attendance, state employee salaries, and health care costs, the model compiles the projected levels of all general fund expenditures and revenues in future fiscal years. The model's current services simulation may then be used as a tool to reflect the changes that may result from program and policy changes as well as amendments in the assumed growth rates.

A representative of the Barents Group noted that the principle benefit of the model is that it provides a systematic, structured method for integrating the best available information on program growth. Much of the data required to be plugged into the model is already produced by agencies, and the executive budget office works jointly with the legislative fiscal office to develop additional information. Fiscal notes showing the projected cost of legislative changes over a five-year period are now incorporated into the model during the legislative process.

The model was praised as a valuable tool in policy decision-making. A plan to increase teacher pay to the national average was cited as a recent example. After the general fund financial model projected the costs of this policy change, a plan to phase in the teacher pay increases over a number of years, coupled with cuts in other programs of $150 million annually, was developed.

The Barents Group has remained a partner in the modeling process, which has evolved over the past six years to add detail while simplifying its usability. The representatives from the state and the consultant advised that development of a model should combine the efforts of both outside assistance and in-house personnel. Other lessons learned from the North Carolina experience include getting input from subject area specialists, emphasizing trends rather than exact numbers, keeping the presentation of model results simple, reexamining model design annually, and tracking model results.

Changes in programs over time have made tracking the accuracy of projections difficult, especially in areas of health care and K-12 education funding. Even in these areas, however, deviations have not been so great as to question the validity of the model. Development of the model over the past six years was estimated to have cost the state between $250,000 and $300,000, plus the in-house costs of operating the model.

Results of Survey on Impact Statements

Staff presented the results of the survey of legislators and legislative staff regarding satisfaction with the legislative impact statement process. By a modest margin, most respondents disagreed with the proposition that the impact statement preparation and distribution processes are satisfactory. Areas of the greatest dissatisfaction involved the timeliness and availability of statements. Most of the criticisms noted in respondents' comments focused on timeliness, lack of objectivity, the quality of analysis, and the inaccessibility of statements.

Draft Recommendations

The impact statement survey did not reveal a consensus for specific changes in the current process. The commission is considering a recommendation that the current impact statement process, which is currently administered pursuant to executive order, be codified.

Other recommendations of the commission under consideration include (i) continuing its work for a second year to examine the feasibility of implementing a legislative expenditure model, (ii) providing the General Assembly with access to agency six-year expenditure needs currently prepared under § 2.1-394 B, and (iii) fixing several technical provisions in planning and budgeting statutes. An outline of an interim report and draft legislation to implement these recommendations will be forwarded to members of the commission and the advisory committee for comments.

The Honorable Joseph V. Gartlan, Jr., Chairman
Legislative Services contact: Franklin D. Munyan