Evaluating Whether Fiscal Notes Include Accurate and Reasonable Estimates

Introduction

Senator Larry Alley requested this audit, which the Legislative Post Audit Committee authorized at its April 25, 2023 meeting.

Objectives, Scope, & Methodology

Our audit objective was to answer the following question:

1. Do agencies provide accurate and reasonable estimates in fiscal notes for potential legislation?

We reviewed 20 bills introduced during the 2018-2022 legislative sessions to answer the question. We based this selection on things like whether the bills were amended, the agencies involved, and the size of the bills’ estimated fiscal effects. For the 10 bills that were enacted into law, we compared their estimated fiscal effects to their actual fiscal effects to determine how closely they aligned. To do this, we talked to agency officials, reviewed fiscal note documentation including agencies’ estimates, and reviewed accounting data. We reviewed data for each year included in the fiscal note. Some fiscal notes included a single fiscal year, but others included multiple years.

For the 10 bills that were never enacted into law, we compared agencies’ processes and controls for creating their fiscal estimates to a handful of federal best practices to determine their reasonableness. To do this, we talked to agency officials and reviewed fiscal note documentation including agencies’ estimates. We also reviewed the Division of the Budget’s processes for coordinating with agencies and accurately reflecting their fiscal estimates in issued fiscal notes.

For both types of work, we excluded fiscal notes from any legislative agencies, including Legislative Post Audit. Including them would’ve impacted our independence.

We also reviewed a couple national best practices for creating fiscal notes and talked to officials from 3 other states to determine how they follow them.

More specific details about the scope of our work and the methods we used are included throughout the report as appropriate.

Important Disclosures

We conducted this performance audit in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. Overall, we believe the evidence obtained provides a reasonable basis for our findings and conclusions based on those audit objectives.

Audit standards require us to report our work on internal controls relevant to our audit objectives. They also require us to report deficiencies we identified through this work. In this audit, we evaluated agencies’ controls for creating reasonable fiscal estimates. We identified several design and implementation deficiencies, as detailed later in the report.

Our audit reports and podcasts are available on our website (www.kslpa.org).

Changes made to bills after the fiscal notes were submitted resulted in most of the inaccuracies we saw, but a few fiscal notes were unreasonable because of agencies’ methods.

Background

Statute requires the Division of the Budget (Budget) to provide fiscal notes for original bills but outlines only a few requirements for them.

• Fiscal notes are documents Budget creates and distributes to legislators for each bill introduced in the Legislature. They show the estimated fiscal effects a bill would have if it were enacted into law. This includes changes in revenues, expenditures, and full-time-equivalent staff positions. Appendix B shows an example fiscal note.

• Statute (K.S.A. 75-3715a) requires Budget to produce fiscal notes for any introduced bill that would:
  • increase or decrease state, county, city, or school district revenues;
  • make or alter state appropriations or fiscal liabilities; or
  • impose county, city, or school district functions or responsibilities that would increase their expenditures or fiscal liabilities.

• Statute further requires each fiscal note to include a specific dollar value estimate of the bill’s immediate effects and a statement of the bill’s long-term effects (although long term isn’t defined). If producing immediate or long-term estimates is impossible, the note must state why. 

• Finally, Budget must provide fiscal notes to the relevant legislative committee within 7 days of a bill’s first reading. Budget officials said they aim to provide fiscal notes by the bill’s hearing date if this is fewer than 7 days after the first reading. Some agency officials told us this often means fiscal notes must be completed in just a couple days.

Affected agencies create fiscal estimates and send them to Budget, which creates the fiscal note for each bill.

• When a bill is introduced, Budget requests fiscal estimates from the agencies the bill would impact if it were to pass. Budget generally requires agencies to use a specific template and expects agencies to provide documentation supporting their estimates. Appendix C shows an example of a completed template.
  • Budget’s template has a table for agencies to estimate the dollar values of the bill’s immediate fiscal effects. It covers the fiscal year the bill is introduced (i.e., the then-current fiscal year) and the following fiscal year. Agencies are to note any changes to revenue and expenditures for the state general fund, fee funds, and federal funds. It also has a place to note any changes to full-time-equivalent staff positions.
  • The template also requires agencies to provide a statement of the bill’s estimated long-term effects. These should cover the years beyond those in the table, but the template doesn’t specify how many years to include. The template doesn’t require specific dollar values for long-term effects.
  • Finally, Budget’s template has a section for the agency’s interpretation of the bill as well as the assumptions underpinning the agency’s estimate. These might include things like inflation assumptions, staff salary and benefit information, or whether costs and revenues are one-time or ongoing. We saw that a couple agencies don’t use Budget’s template. They use agency-created templates instead, which Budget officials accepted.

• Agencies decide how to approach their estimation processes, and the agencies we reviewed did this differently. For instance, the agency’s response may be completed or reviewed by program staff, legal staff, or finance staff. It may lay out many detailed assumptions and calculations or provide bottom-line totals and almost no methodological information.

• For bills affecting counties, cities, or school districts, statute (K.S.A. 75-3715a) directs Budget to consult with the relevant association as appropriate (e.g., Kansas Association of Counties, League of Kansas Municipalities). These estimates generally don’t include dollar values because there would be too many entities involved to meet statute’s 7-day deadline.  

Budget staff conduct a high-level review to determine whether agencies’ estimates seem reasonable.

• The assigned Budget analyst reviews agencies’ completed estimation templates before incorporating this information into the fiscal note. Budget analysts’ reviews are high level because they must produce fiscal notes for hundreds of bills on a quick turnaround each session. Budget has 10 analysts, and they issued about 600 fiscal notes each session during 2018-2022.

• Budget analysts may exclude some of the details from agencies’ estimates in the final fiscal note. For example, agencies may describe the detailed assumptions they used to generate their estimates in the template they complete and send to Budget. Budget analysts may use that information to determine the validity of an agency’s estimate but not include those details in the fiscal note itself. Budget also doesn’t attach agencies’ completed estimation templates to fiscal notes.

Statute only requires fiscal notes when bills are introduced and doesn’t require updates to reflect subsequent changes made as bills move through the legislative process.

• Fiscal notes provide an initial estimate of a bill’s impact as originally written. Figure 1 shows the opportunities for bills to be amended as they move through the legislative process. As the figure shows, bills can be amended at several points in the process. Statute doesn’t require Budget to update fiscal notes in response to bill amendments. As such, fiscal notes often don’t reflect the impact of the amended bill.

• Budget officials said they provide updated fiscal notes to legislators who request them, but this is rare. They told us legislators only request updated fiscal notes about 6-7 times per session. Budget sends these updated fiscal notes directly to the requesting legislator. They don’t distribute them more widely, and they don’t officially replace the original fiscal note.

• Budget also issues corrected fiscal notes if they catch a mistake in an issued fiscal note. This is also rare. For instance, Budget issued only 4 corrected fiscal notes during the 2023 session.

• Finally, updated fiscal estimates sometimes appear outside the fiscal note process. For instance, agencies may provide updated estimates through legislative testimony. Legislative Research officials said they may distribute agencies’ updated estimates through supplemental notes or the omnibus memorandum, but Legislative Research staff don’t review the updates for accuracy.

We reviewed 20 fiscal notes for bills that were enacted and bills that died.

• To determine whether agencies created reasonable fiscal estimates, we looked at 10 bills that were enacted and became law and 10 bills that died and were never enacted into law. We detail each bill and fiscal note and the results of our reviews in Appendix D.
  • To determine accuracy, we selected 10 enacted bills from the 2018-2019 legislative sessions and compared their estimated fiscal effects to their actual fiscal effects. We selected older bills to allow time for the actual effects to become clear. We reviewed data for each year included in the fiscal note, which varied across bills. Since the future is inherently unknowable, we generally considered estimates that were within 20% of actual effects to be accurate. If we found a greater difference, we followed up to understand why.
  • To determine reasonableness, we selected 10 bills that died from the 2021-2022 sessions and reviewed whether agencies followed reasonable processes to estimate their fiscal effects. We selected more recent bills to review more recent processes. We followed up with the agencies involved to understand any shortcomings we saw.

• Both bill selections varied on things like the number of affected agencies, whether the agencies frequently created fiscal estimates, and whether the bill affected local governments. They also varied on things like whether the bill created a new or modified an existing program, the size of the fiscal estimate, and whether subsequent amendments changed the bill. The 20 bills we selected included estimates from 16 state agencies and associations, as shown in Appendix D.

• Our selections of agencies and bills provide good cross-sections of each. But both selections are small relative to their populations, and each bill is unique. As such, our conclusions can’t be generalized to other agencies or bills, and there may be problems with fiscal notes we didn’t review. But we think our analyses provide useful information about why estimated fiscal effects and actual fiscal effects may not always align.

Accuracy of Fiscal Notes for Bills Enacted into Law

7 of 10 fiscal notes for enacted bills differed significantly from their actual fiscal effects.

• To determine the accuracy of fiscal notes, we selected bills that passed from several years prior. We then compared the original fiscal estimate to the actual fiscal effects since enactment. In general, we determined fiscal notes were accurate if they were within 20% or a few thousand dollars because precisely predicting the future is impossible. We thought close estimates were more realistic to expect.

• We reviewed fiscal notes for 10 bills that were enacted into law during 2018-2019. Figure 2 shows the bills and whether and how much their actual fiscal effects differed from the original fiscal note. As the figure shows, 7 bills’ fiscal estimates were significantly different from their actual costs. Their actual costs ranged from 279% less than the fiscal note to 133% more than the fiscal note.

• Fiscal notes can be inaccurate for a variety of reasons. Some may be attributable to problems with the original fiscal note. For example, a fiscal note that uses the wrong data source or poor assumptions. But some reasons are unavoidable because of the inherent difficulty involved with predicting the future. For example, a fiscal note that requires estimating the number of Kansans a new program will serve.

• We discuss the 7 fiscal notes that differed from their actual fiscal effects later in this report.

• The remaining 3 fiscal notes we reviewed were accurate. As shown in Figure 2, their actual fiscal effects were within 20% or a few thousand dollars of the fiscal note. These bills had narrow purposes and agencies had relevant historical data to use for the fiscal note. For example, 2018 SB 256 designated a memorial highway section and required the Department of Transportation to install 2 signs. The bill was straightforward, and the agency had prior experience installing highway signs.

• Although these 3 fiscal notes were accurate, that doesn’t mean the agencies did everything perfectly. They may have incorrectly predicted the timing of when revenue would come in, for instance. But these mistakes didn’t affect the fiscal notes’ fundamental accuracy.

Most of the differences between estimated fiscal effects and actual fiscal effects were caused by bill amendments or other changes made after the fiscal notes were submitted.

• As Figure 2 shows, 4 bills were substantively amended after Budget issued the fiscal notes. The amendment was the main reason the bills’ fiscal estimates were so different from their actual fiscal effects. For example:
  • 2018 HB 2579 allowed wrongfully imprisoned people to seek damages against the state. Bill amendments lowered the monetary damages claimants could receive each year and generally limited attorney’s fee compensation to a total of $25,000. The bill hadn’t specified a limit before. The amendments also clarified the other kinds of damages claimants could receive, such as tuition assistance and health insurance. These changes resulted in the bill’s costs being lower than anticipated.  
  • 2019 SB 187 increased fees for certain overweight or oversized vehicle permits. The effective date of the original bill left a 6-month gap when the Department of Transportation wouldn’t have received any permit fees. Bill amendments closed this gap. This resulted in the department’s revenues being higher than it had originally estimated.

• As Figure 2 also shows, 2 other fiscal notes didn’t align with actual fiscal effects because agencies implemented the bill using a different process than they anticipated or fewer Kansans participated in the program than agencies predicted.
  • 2019 SB 211 created a Youth Suicide Prevention Coordinator position in the Office of the Attorney General. The position was originally intended to be full-time, and the fiscal estimate reflected this. But after the bill was enacted and the position was created, the Office of the Attorney General initially filled it with a part-time employee, which cost much less than a full-time employee.
  • 2019 HB 2044 created a new income tax credit for purchases from qualified vendors that employ Kansans with disabilities. The Department of Revenue estimated the new credit would reduce state tax revenues by $1.7 million annually. This was based on taxpayers making $11.3 million in qualifying purchases. But agency officials said they didn’t have historical data useful for predicting how many taxpayers might claim the credit. Fewer than 5 claimed it, so the credit’s revenue effects were less than estimated. This was the primary reason the estimate was off. But a couple other things affected it, too, like the Department of Revenue’s IT services contract being canceled. This reduced the credit’s administrative costs.

• Although these fiscal notes didn’t accurately reflect the bills’ actual costs, we don’t think they indicate a systemic problem with agencies’ estimates. That’s because changes that happened after the fiscal notes were written were the main reason the estimates were inaccurate. Neither agencies nor Budget could predict subsequent bill amendments, but this problem could be addressed if the Legislature required fiscal notes to be updated. Inaccuracies from things like program participation are likely unavoidable.

Only 1 fiscal note was inaccurate because of the agency’s methodology.

• 2018 HB 2511 required renewal of commercial driver’s licenses every 5 years rather than every 4 years. The Department of Revenue estimated the impact of this change over 20 years and divided those costs evenly across each year.  In reality, the costs associated with this change appear to be larger in earlier years than in later years. The department’s method of averaging costs was the primary reason its estimate was significantly different than the actual cost in its first year (the only year for which actual costs were available).

• Although the department’s method likely was reasonable for approximating total long-term costs, it didn’t accurately anticipate the immediate annual costs.

Our 2010 audit found that fiscal notes generally differed from their bills’ actual effects for these same reasons.

• LPA conducted a similar audit of fiscal notes in 2010. As part of that work, we compared bills’ estimated fiscal effects to their actual fiscal effects. We also determined the reasons for any differences we saw.

• That audit and this one reached similar conclusions. Our 2010 audit reviewed 8 fiscal notes for enacted bills. Of those:
  • 3 fiscal notes differed because the Legislature or Governor changed the scope or funding for the program after Budget issued the fiscal note.
  • 3 fiscal notes differed because of unpredictable program participation. For 1 of these 3, the agency also failed to account for all the information it should’ve known.
  • 2 fiscal notes differed because the agency changed the program’s scope or funding.

Most other states update fiscal notes in response to bill amendments, unlike Kansas.

• According to a 2015 report by the nonpartisan Center on Budget and Policy Priorities, 26 states updated fiscal notes to reflect changes from bill amendments. Kansas currently doesn’t do this even though there are multiple opportunities for bills to be amended during the legislative process. None of the agencies we talked to maintained statistics on how many bills are amended in Kansas each session, but our review of legislative bills suggests it’s very common. Amending bills without updating the corresponding fiscal notes makes the fiscal notes less accurate and helpful.

• We talked to officials from 2 states that routinely update fiscal notes when bills are amended to understand how it’s feasible for them to do so. Both Nebraska and North Dakota officials said they limit when fiscal notes are updated to certain points in the legislative process. They said they update fiscal notes only when bill amendments pass to the next stage of debate.
  • Nebraska’s unicameral legislature also naturally limits the number of amendment opportunities for any given bill. Nebraska’s fiscal office has 9 analysts, which is about the same as Budget. Nebraska officials said they produced fiscal notes for about 800 bills during the 2023 session and updated about 300 to 400 of them.
  • North Dakota only requires original fiscal notes and fiscal note updates when the estimated fiscal effect is at least $5,000. This likely limits the number of fiscal notes they issue in the first place and therefore how many revisions they must make. North Dakota officials said their 7 analysts produced about 250 fiscal notes during the 2023 session. They updated about 150 of them.

• Finally, Nebraska and North Dakota officials said they follow tighter timelines for fiscal note updates. Nebraska told us they had only 2-3 days to revise fiscal notes in response to bill amendments, while North Dakota told us they had 1 day. This would likely prevent revisions from holding up the legislative process too much.

Reasonableness of Fiscal Notes for Bills That Died

We reviewed whether agencies followed certain best practices to help ensure fiscal notes had reasonable estimates.

• We reviewed fiscal notes for 10 bills that died during the legislative process. Because the bills weren’t enacted into law, we couldn’t compare their estimated fiscal effects to their actual fiscal effects. Instead, we reviewed agencies’ processes for creating the fiscal notes to determine if they included the U.S. Government Accountability Office’s best practices for developing fiscal estimates.
  • We focused on agencies’ processes because they’re largely responsible for creating the estimates that Budget includes in the fiscal notes. We looked at whether agencies used the following 6 practices: (1) ensuring they understood the bill language and fiscal effects, (2) identifying assumptions they would need to make, (3) considering the effects of alternative assumptions, (4) assessing whether they had accurate and complete data, (5) documenting their methodology and assumptions, and (6) ensuring estimates were reviewed by a second person.
  • For fiscal notes affecting more than 1 agency, we also looked at Budget’s process for coordinating and validating the agencies’ work.

• We asked agency officials to describe how they generally create fiscal notes. We then reviewed documentation to determine whether they followed the processes they described on the fiscal notes we selected. Creating fiscal notes is a low-level agency process, so we didn’t expect to see agencies’ processes detailed in policy. And we only reviewed a few fiscal notes for each agency, which may be a small fraction of what they create each session. As such, our work can’t be generalized to all the affected agencies’ work on fiscal notes.

3 of the 10 fiscal notes we reviewed didn’t appear reasonable because they didn’t include complete or correct information. 

• Figure 3 shows the 10 bills we reviewed and whether we thought their fiscal notes included reasonable estimates based on the processes agencies used to create the notes. As the figure shows, the affected agencies and Budget created 7 fiscal notes that appeared to be reasonable and 3 that didn’t appear reasonable. For the 7 reasonable fiscal notes, agencies may not have perfectly documented their processes or had a couple processes that were inadequate. But they generally followed the best practices and controls they described.

• For example, 2021 SB 272 would’ve required the KBI to conduct urinalysis tests for certain controlled substances as part of all sexual assault kits. We reviewed KBI officials’ estimates for how much it would cost for their lab to do the tests. We thought they identified all important fiscal effects, used accurate historical data and equipment costs, and applied appropriate assumptions to come up with their estimate.

• The remaining 3 fiscal notes didn’t appear to be reasonable because they didn’t include complete or correct information, as discussed below.

2 fiscal notes appeared to be unreasonable because the agencies used questionable data and assumptions.

• 2022 SB 469 would have allowed people with felony drug convictions to receive food assistance. The Department for Children and Families estimated how much it would cost to process the new applicants. But we weren’t sure their estimate was reasonable because they used questionable assumptions to estimate the number of new food assistance cases the bill would create.
  • The department estimated the bill would increase the agency’s food assistance caseload by fewer than 50 cases per month, which is likely too low. They based this on the number of applicants and recipients historically rejected each month due to drug convictions. But department data shows that if those who were rejected during 2019-2020 reapplied within 6 months of the bill’s passage, this would be about 750 people, or 125 per month. And many people who don’t appear in the data may apply, too. Some people likely didn’t apply for assistance before because they knew their convictions disqualified them.
  • Department officials told us the estimated fiscal effect likely would’ve remained $0 even if they had used different assumptions because the increased caseload could have been absorbed by existing staff. We don’t know whether this is true. 

• 2022 SB 516 would’ve created a sales tax exemption for personal hygiene products like soap, diapers, and feminine hygiene products. The Department of Revenue estimated the bill would decrease state tax revenues by $19.7-$22.3 million annually during fiscal years 2023-2025. But we weren’t sure this was accurate because department officials relied on several assumptions we didn’t think were reasonable.
  • The department calculated Kansans’ personal care spending on tax exempt items would be about $70-100 per year, which is likely too low. For example, according to some national estimates we reviewed, women spend about an average of $240 per year on feminine hygiene products alone. Because the bill also would have exempted many other items, it seems likely people would spend much more on tax exempt items than the department assumed.
  • To estimate diaper costs, the department assumed Kansas births would be comparable to the average annual number during 2017-2020, but this is likely too high. Births generally have been declining each year since 2014. The department’s method overestimated births by about 1,000 for 2021 and didn’t reflect any decreases after then. This means they may be overestimating diaper costs significantly each year by 2023-2025, the years for which they’re calculating.

• These bills demonstrate the importance of agencies using data and assumptions that are complete, accurate, and well thought out. Overall, the 10 fiscal notes we reviewed revealed agencies generally followed the best practices. We saw only a few repeated instances of agencies not following good processes. For example, of the 6 national best practices we reviewed, we only saw repeated issues with 2 (data validation and documentation of agency methods and assumptions). The problems we saw with agencies’ processes seem to be on a case-by-case basis and not systemic.

1 fiscal note appeared to be unreasonable because Budget didn’t include a potential large cost that an agency had identified.

• 2021 HB 2241 would’ve required all health insurers to cover certain diagnostic breast examinations. The bill affected the Department of Administration and the Kansas Insurance Department. The fiscal note accurately reflected the Department of Administration’s estimate of the cost to that agency.

• The fiscal note didn’t include all the potential costs the Kansas Insurance Department identified. The Kansas Insurance Department said the bill could trigger a large state cost. The diagnostic breast examination coverage included in the bill isn’t required by federal law, which means the state would have been required to help health insurers cover the cost of such procedures. A similar fiscal note in Nebraska estimated a $6 million annual cost. Budget told us leaving this information out was likely just a mistake.

• Overall, we don’t think this indicates a major problem with the fiscal note process. This fiscal note had a large error. But we saw other cases where Budget identified and corrected agency mistakes. As such, we can’t say Budget’s mistake on this fiscal note is indicative of a systemic problem.

We didn’t see evidence Budget officials always coordinated with agencies on fiscal notes like we would’ve expected.

• Agencies are the program experts, so they create the dollar values used in fiscal notes. But Budget is responsible for the finalized fiscal notes issued to legislators. This includes checking agencies’ work and accurately translating the agencies’ responses into the fiscal notes. It also includes coordinating agencies when more than 1 is involved, to ensure they understand the bill the same way and use compatible methods.

• 4 of the fiscal notes for bills that died that we reviewed involved multiple agencies and therefore required Budget to coordinate. We found a problem with Budget’s coordination in 1 of these 4 instances. In 2 others, we couldn’t tell whether Budget communicated with the Kansas Association of Counties or League of Kansas Municipalities like we would’ve expected.
  • 2022 HB 2685 would have required all Kansas students to show grade level proficiency on new standardized tests. Budget asked the Kansas State Department of Education to estimate the cost to develop and oversee the testing. But they didn’t coordinate with the Kansas Association of School Boards. Statute (K.S.A. 75-3715a) requires this when appropriate. We think they should’ve in this case because the bill could affect school districts financially, especially if students fail and are held back. Budget said they didn’t have a process for consulting with the association and thought the State Department of Education had the needed information.
  • 2022 HB 2482 would’ve created a sales tax exemption for car rentals. The Department of Revenue estimated the bill would decrease state revenues. But the Kansas Association of Counties and League of Kansas Municipalities misinterpreted the bill and said it would increase local revenues. Budget corrected this and included the right information in the fiscal note. But we didn’t see evidence Budget communicated this change to the associations, who also didn’t recall this happening. Budget said this usually happens informally via phone.
  • Similarly, 2021 SB 70 would’ve made permanent a sales tax exemption for new vehicle cash rebates and created an exemption for manufacturer coupons. The Department of Revenue estimated the bill would decrease state revenues. But the Kansas Association of Counties and League of Kansas Municipalities said it would have a negligible effect. Neither association noted the new manufacturer coupon exemption. Budget corrected this and included the right information in the fiscal note. But we again didn’t see evidence Budget worked with the associations on this.

Other Finding

Two nearby states have more robust review processes for fiscal notes.

• During our work, we spoke with officials from two other states who described more robust review processes. Iowa and Nebraska officials described scrutinizing agencies’ work more closely than Budget does. Iowa officials said they use data from the state budget system, the National Conference of State Legislatures, and other states to double-check agencies’ figures. But Iowa’s fiscal office has 17 analysts, nearly twice the number Budget has.

• Nebraska officials said they conduct their own analysis to include in each note, indicating whether they agree with the agency. They said they have 9 analysts, who work on roughly similar numbers of fiscal notes as Budget does.

• As noted previously, Kansas Budget officials said they conduct a high-level review of agencies’ responses and frequently defer to agencies’ expertise. Budget officials also described conducting a second review to double-check fiscal notes, but this review focuses on typos. Budget officials told us the second reader doesn’t check the agency’s template or other supporting documentation against the fiscal note.

Conclusion

Our work suggests the cost estimates in fiscal notes frequently don’t match their actual costs. In both this audit and our 2010 audit, we saw that most fiscal notes we reviewed didn’t match their eventual costs after the bill was enacted. However, most of the differences are likely unavoidable under the current process. Most importantly, that’s because statute doesn’t require fiscal notes to be updated when bills are amended. As a result, the cost estimates do not keep pace with changes to bills. Other factors include the tight timeframes to make estimates and that predictions like future participation in new programs are inherently unknowable. For the most part, our review showed agencies are making good efforts to produce accurate and reasonable estimates, and the problems we saw were not pervasive. Going forward, improvements to the process would require large changes. For example, requiring fiscal notes to be updated when bills are amended could help. However, the Legislature would need to consider some parameters to ensure such a process is feasible and adequately resourced. And even then, it’s likely there will always be some differences because it’s difficult to predict the future.

Recommendations

We didn’t make any recommendations for this audit.

Agency Response

On November 9, 2023, we provided the draft audit report to the Division of the Budget, Board of Nursing, Kansas Association of Counties, Kansas Bureau of Investigation, Kansas Insurance Department, League of Kansas Municipalities, Office of the Attorney General, Office of Judicial Administration, and the Departments of Administration, Aging and Disability Services, Children and Families, Commerce, Corrections, Education, Health and Environment, Revenue, and Transportation.

Because we didn’t have any recommendations for these agencies, their responses were optional. The Division of the Budget, Board of Nursing, Department of Revenue, and League of Kansas Municipalities submitted responses. Agency officials generally agreed with our findings and conclusion.

Division of the Budget Response

Thank you for the opportunity to provide an official response to your audit titled: Evaluating Whether Fiscal Notes Include Accurate and Reasonable Estimates (the audit).  The Division of the Budget (DOB) takes our statutory responsibility to provide accurate and impartial fiscal notes very seriously, and we take great pride in the work that we do.  We also are always seeking opportunities to improve our process, and an audit such as this provides an opportunity for self-reflection and for potential procedural updates to consider.

While the audit covered a lot of ground, and provided a comprehensive review of the process, I feel that it is important to highlight a few areas of the audit that need greater emphasis to provide the proper context for policy makers to truly understand the results.  Without this context, some headlines could lead to a misinterpretation of the results.  This official response will be focused on two main points: 1) the lack of relationship between a fiscal note for an original bill and the results of an amended bill, and 2) DOB’s track record and process for coordinating across agencies.

1) The moment a bill is amended, the original fiscal note can no longer be tied to the newly amended bill, thus rendering a comparison of the estimated fiscal impact for an original bill to the actual impact of an amended bill invalid.

The audit states “7 of 10 fiscal notes for enacted bills differed significantly from their actual fiscal effects.”   What is missing from the headline is the fact that four (4) of these fiscal notes are associated with bills that saw significant amendments after the fiscal note was submitted.  The audit does provide this critical context in other areas of the audit, but its lack of inclusion in this headline could lead readers to an inaccurate conclusion of the results of the audit.  As stated in point 1 above, once a bill is amended, there ceases to exist any relationship between the fiscal note and the newly amended bill.  This is important context, because absent that awareness, the headline would suggest that 70% of the bills reviewed in the audit had problems with the original fiscal note, which is not the case, as is discussed in other sections of the audit. 

The audit provides a useful illustration of how often a bill could be amended throughout the legislative process (Figure 1).  If a bill is amended at any point in this process, Legislators should have no expectation of a valid relationship between the original fiscal note and the newly amended bill.  As such, the original fiscal note should not be used as a comparison to the actual impacts of an amended bill.  As is stated in the audit, several agencies produce updated information for many of the amendments that are made, but these are inherently outside of the original fiscal note.  Additionally, Kansas Legislative Research Department (KLRD) often provides updated information on amended bills through Supplemental Notes, which are sometimes shared on the Legislature’s website as bills move through the process.  It is the totality of this information that provides policy makers with a more representative picture of fiscal impacts for amended bills.

Of the remaining three (3) fiscal notes that the audit identifies as inaccurate, all three appear to have provided reasonable estimates for the Legislature to consider when debating the underlying bills, once the details of the discrepancy are understood.  One fiscal note stated that the agency would need one full FTE, but upon implementation of the program, the agency made a decision to instead fill that position with a part time employee.  One fiscal note stated that the agency lacked the necessary historical information to provide an accurate estimate, and this was clearly stated in the fiscal note.  The agency chose to report the highest possible fiscal impact of the legislation (again, clearly identified in the fiscal note), which did not come to fruition, based on less than full participation in the program upon implementation.  In the final instance, the audit identified that the fiscal note provided an accurate representation of the lifetime cost of the bill, but the timing of the cost did not align with what was reported in the fiscal note.

2) DOB has demonstrated a strong history of effective coordination across agencies for many years; some of this coordination occurs via phone, which does not mean that DOB does not coordinate with agencies.

The audit states “We didn’t see evidence Budget officials always coordinated with agencies on fiscal notes like we would’ve expected.”  Two specific examples cited in the audit are from bills in which the agency incorrectly interpreted a bill’s effect when sending their fiscal impact to DOB, and DOB fixed the error prior to submitting the official fiscal note, thus producing an accurate fiscal note.  The audit notes that there is no evidence that DOB contacted the agency prior to fixing the fiscal note.  While there is admittedly no email trail to substantiate DOB’s position, we do have a policy that we will contact agencies any time we are making a substantive change to the fiscal impact that they send to us.  Many times, this is most easily completed with a phone call, which would not produce any historical evidence to rely upon.  However, the lack of such a trail does not necessarily indicate a lack of coordination between agencies on DOB’s part.  Rather, DOB has consistently demonstrated coordination between agencies, as is evidenced by the collection of necessary data from impacted agencies, and ensuring consistent reporting, as happened with these two bills.  One takeaway that DOB has from this portion of the audit is to determine how we can better document our discussions in situations such as this, to ensure we have contemporaneous records of our conversations.

In closing, I want to thank you and your team for the professional and thorough manner in which you conducted this audit.  You and your team provided ample communication, and you allowed for thoughtful and deep conversation during the process.  As was stated in the opening, DOB prides itself on producing accurate, timely, and impartial fiscal notes for the Legislature to review when considering legislation.  This audit produced an opportunity for us to ensure we have the proper infrastructure and policies in place to best serve the citizens of Kansas.  We will continue to have internal discussions to determine what, if any, changes will be made to our current practices moving forward.

Board of Nursing Response

The Kansas State Board of Nursing concurs with findings of this Legislative Post Audit, Evaluating Whether Fiscal Notes Include Accurate and Reasonable Estimates, Report number: R23-013. We agree there are opportunities to improve the process, so the fiscal notes are more accurate. The fiscal note submitted for HB 2496 did not reflect the amendment that included an increase in licensure fees to assist in offsetting the revenue loss and the significant grant from NCSBN for computer system upgrades, staff training and communication to licensees.

Department of Revenue Response

Thank you for the opportunity to review and comment on your completed performance audit, Evaluating Whether Fiscal Notes Include Accurate and Reasonable Estimates. The Kansas Department of Revenue (Department) prepares estimates for proposed legislation and seeks to provide valuable information to the Division of Budget to incorporate into the official fiscal note prepared for the Kansas Legislature.

As provided in the audit, K.S.A. 75-3715a requires fiscal notes for original bills only. However, the Department does provide updated estimates on applicable fiscal notes for each amendment to a bill as it proceeds through the process. These updated estimates are shared with the Division of Budget and the Kansas Legislative Research Department. During the 2023 legislative session, the Department prepared 320 fiscal note estimates for proposed legislation and reviewed almost 100 additional pieces of proposed legislation to determine any relevant estimated fiscal impacts and Department implications.

As a matter of practice, the Department does make every effort to use adequate data sources, sound methodologies and accurate calculations when estimating impacts for proposed legislation. The time consuming tasks of reviewing the often-complex, multifaceted proposed legislation, considering administrative impacts or mechanics, compiling applicable information from the various Department operating units, preparing the estimated fiscal impact, and reviewing the estimated fiscal impact are subject to very tight time constraints. The Department makes it an absolute priority to provide detailed information in a timely manner after having exercised proper due diligence and having followed accepted estimating protocols.

Thank you to the Legislative Post Audit staff for their work on this performance audit and thank you for the opportunity to participate in this review.

League of Kansas Municipalities Response

Thank you for the opportunity to provide feedback on the Kansas Legislative Division of Post Audit’s (LPA) Evaluating Whether Fiscal Notes Include Accurate and Reasonable Estimates report. The League of Kansas Municipalities (League) provided information to the Legislative Post Audit team on August 8, 2023 and participated in an interview on August 24, 2023.

The League reviewed the draft report (Evaluating Whether Fiscal Notes Include Accurate and Reasonable Estimates) and provided some key points to the LPA staff, which are now addressed in the final report. The League would, however, like to emphasize three important matters that speak to some of the larger issues at play in the legislative process.

• In two instances, the report indicates the League’s fiscal note was either incorrect or misinterpreted. That assessment may be correct, but—regardless—we think it is important for the audit to acknowledge that in both cases we did not hear from the Budget Office to inquire about the discrepancy, ask for more information, or provide us information regarding changes in the legislation that may have affected the analysis.

• The average requested turnaround time for the League is 48 hours or less from request to wanting a completed fiscal note. It is not uncommon for a request to come in on a Friday, with the deadline being Sunday. Most offices, including ours, do not have our full staff working on weekends. These turnaround times provide little time for our organization to gather the most comprehensive information relating to a fiscal-note inquiry. It is this element (the rushed nature of the legislative process) that leads to ill-considered laws and policies.
  • To this point, the current draft of the audit minimizes the time factor and the short timeframe in which agencies are asked to provide information. It mentions that a fiscal note is required within 7 days of a bill reading. Yet it does not address the reality that the average time for agencies is significantly less than the one-week interval.
  • As we stated at the beginning, the turnaround time is the League’s top concern, and improving on the research processes will only help result in better information and improved legislative results.

• The fact that fiscal notes do not receive updates after major changes are made to legislation is major flaw in the legislative process. We understand that changes occur rapidly and often, so the process itself is fraught with challenges. We would like to mention that the failure to include a process for updates—particularly for large pieces of legislation—can negatively affect cities and the state as a whole because without that updated information, the resulting picture is incomplete and fails to capture the overall budget implications before a bill goes to final action.

Appendix A – Cited References

This appendix lists the major publications we relied on for this report.

1. Better Cost Estimates, Better Budgets (November, 2015). Center on Budget and Policy Priorities.
2. Cost Estimating and Assessment Guide (March, 2020). U.S. Government Accountability Office.
3. Standards for Internal Control in the Federal Government (September, 2014). U.S. Government Accountability Office.

Appendix B – Example Fiscal Note

This appendix includes an example of a fiscal note the Division of the Budget issued during a recent legislative session. It reflects the agency response in Appendix C.

Appendix C – Example Agency Response

This appendix includes an example of a completed agency response template the Division of the Budget received during a recent legislative session. It informed the fiscal note in Appendix B.

Appendix D – Reviewed Fiscal Note Details

This appendix details the 20 bills we selected for review in this audit and the results of our review. It shows both the 10 bills enacted into law and the 10 bills that died.