Fiscal Guardrails: Navigating Volatility, Finding Stability

01.31.2025
Issues & Policies

    Understanding Connecticut’s fiscal guardrails, particularly the volatility cap, requires first grasping why certain revenue sources are considered “volatile” and how this volatility creates unique challenges for state budgeting.

    The state’s most significant volatile revenue sources—estimated and final payments from the personal income tax and the pass-through entity tax—exhibit substantial year-to-year fluctuations that make them particularly difficult to predict and dangerous to rely upon for ongoing expenses.

    Historical data shows these sources have volatility of approximately 19.9% in year-over-year changes, meaning that in any given year, these revenues have been likely to swing up or down by up to one-fifth of their value.

    Compare this to simple withholding revenues from worker tax returns, which see a year-to-year volatility of +/- 3.34%.   

    These dramatic fluctuations aren’t random—they’re tied closely to economic factors, particularly financial market performance.

    A significant portion of these revenues comes from capital gains realizations, which are heavily influenced by stock market returns.

    Market Performance

    The connection to market performance makes these revenues both more volatile and more difficult to forecast than other state revenue sources.

    For example, during the period since the fiscal guardrails were implemented, the S&P 500 has seen three years with returns exceeding 25%—2019, 2021, and 2024—representing an extraordinary concentration of high-performing market years.

    This level of market performance is historically unusual; in the period from 2001 to 2024, only two other years saw returns of this magnitude (2003, 2013).

    With that in mind, looking ahead it may not be prudent for policymakers to consider the past few years as a “new normal” for volatile revenues. 

    It may not be prudent for policymakers to consider the past few years as a “new normal” for volatile revenues.

    The challenge of managing these volatile revenue sources extends beyond their relationship to the markets.

    Unlike more stable revenue sources such as sales tax or withholding from regular wages, these sources can see sudden, dramatic shifts based on changes in federal policy or taxpayer behavior.

    The federal Tax Cuts and Jobs Act of 2017 provides a telling example—its provisions encouraging repatriation of foreign profits led to a significant one-time surge in collections during fiscal year 2018.

    This is displayed by the significant jump in EF revenues in 2018 (+46.1%), while we would have only expected perhaps a 19.5% increase in volatile revenues based its historical relationship to capital gains.

    This type of policy-driven spike in revenues illustrates why treating these sources differently isn’t just prudent—it’s necessary for maintaining fiscal stability. 

    Pass-Through Entity Tax

    Some have pointed to the introduction of the pass-through entity tax as a volatile revenue source as a possible source of more stable funds.

    However, in the five years that we have data on that tax, the average one-year variance has been 12.9%, but has ranged from a 48.9% gain to an 11.2% decline.

    During the same period, Estimates and Finals has remained volatile as well, seeing one-year average changes at 4%, but a range of +37.1% to -31.5%.

    Overall, these two sources have remained volatile during the fiscal guardrails period, even if they have been at a higher level overall.

    It is common to see near 20% swings in these revenues on a year-to-year basis.

    So where should we expect volatile revenues to go from here?

    By their very nature, it is hard to predict what revenues will be from these sources, but we can use their historical changes to make some estimates.

    For example, as noted before, we know that it is common to see near 20% swings in these revenues on a year-to-year basis, which would equate to $1.06 billion based on fiscal 2025 volatile revenue estimates ($5.3 billion as of Nov. 12, 2024).

    If that adjustment happened to the downside in fiscal 2026, it would leave volatile revenue just above the fiscal 2026 projected volatility cap of $4.095 billion. 

    Ultimately, it’s difficult to know where these revenues will stand in the years to come, however market returns and policy changes have been unique in recent years, which is worth considering when we project revenues in the future. 

    Volatility Cap: Purpose, Design 

    Acknowledging the inherent volatility of Connecticut’s income, lawmakers in 2017 established the state’s volatility cap, which caps how much of the state’s volatile revenues can be budgeted for in a given fiscal year.

    However, Connecticut’s volatility cap serves as more than just a mathematical formula or budgetary constraint—it represents a fundamental shift in how the state approaches fiscal planning in the face of uncertainty.

    While the mechanics of the cap are relatively straightforward, its purpose speaks to deeper lessons learned from decades of state budgeting challenges. 

    Budgeting with Uncertainty 

    The fundamental challenge that the volatility cap addresses is not just that certain revenue sources fluctuate—it’s that these fluctuations create asymmetric risks in the budgeting process.

    When policymakers are presented with revenue projections showing high levels of volatile revenue sources, they face immense pressure to budget based on those projections, even knowing the inherent uncertainty in those numbers. 

    This creates what economists call a “ratchet effect” in state spending. During good years, spending tends to rise to match available revenue.

    However, when revenues decline sharply, it becomes politically and practically difficult to reduce spending by corresponding amounts.

    When revenues decline sharply, it becomes politically and practically difficult to reduce spending by corresponding amounts.

    Connecticut’s experience during the 2008-2010 financial crisis provides a stark illustration of this challenge.

    When revenues declined sharply, the state found itself unable to readily reduce spending to match.

    Instead, Connecticut was forced to take on emergency debt to maintain spending levels and even redirected planned pension contributions to cover operational expenses, further exacerbating the state’s long-term liabilities.

    This asymmetry—finding it easier to spend in good years than to cut in bad years—can lead to structural budget problems that persist long after the initial crisis. 

    Predictably Difficult to Predict 

    Revenue projections rely heavily on recent historical data and current trends. This can create a form of recency bias where unusually strong recent performance (like the market gains of 2019-2024) may lead to overly optimistic projections.

    Connecticut’s current situation illustrates this challenge perfectly—the most recent consensus revenue forecasts from the Office of Policy and Management project essentially flat revenue from volatile streams over the next several years.

    While this conservative projection approach is understandable given the inherent uncertainty, historical patterns suggest these revenues are highly unlikely to remain flat.

    Instead, they will likely deviate significantly either above or below these projections, highlighting the fundamental challenge of relying on such forecasts for budgeting purposes.

    In fact, since the guardrails were implemented for fiscal 2018, November consensus revenue projections for EF and PTE revenues have missed actual revenues by an average of 24.1%, roughly $883 million each year. 

    Fortunately, these were primarily underestimates, however in fiscal 2023 projections overestimated revenues by $525.7 million.

     The political dynamics of the budgeting process tend to favor accepting optimistic revenue projections, especially when recent performance appears to justify such optimism.

    This creates a systematic bias toward treating recent strong performance as the “new normal,” making it tempting to build spending plans around revenue levels that may not be sustainable.

    As recent years show however, treating such projections as reliable carries a certain amount of risk. 

    Practical Function 

    This is where the volatility cap’s design becomes crucial. By establishing a conservative baseline for volatile revenue sources that can be relied upon for budgeting, the cap serves several practical purposes: 

    1. It creates a binding constraint that helps policymakers resist the temptation to budget based on optimistic projections of volatile revenue sources, even when recent performance might seem to justify such optimism. 
    2. It forces the state to treat revenue above the cap as truly extraordinary rather than building it into baseline spending expectations. This helps prevent the ratchet effect where temporary revenue spikes lead to permanent spending increases. 
    3. It provides a mechanism for systematically directing extraordinary revenues toward long-term fiscal priorities like building reserves and paying down unfunded liabilities, rather than leaving such decisions to annual budget negotiations. 

    On the first two points, we already see a benefit when we compare Connecticut’s current fiscal situation with that of many other states.

    Windfall dollars during the post-pandemic period created significant booms in state budgets.

    Windfall dollars during the post-pandemic period, driven by a combination of inflation and fiscal stimulus, created significant booms in state budgets.

    However, today many now face structural shortfalls in their budget due to ebbing COVID related windfalls coming up against expanded permanent spending.

    In a January 2024 article, Pew Research pointed out that states across the country are projecting deficits going forward after years of surplus, including Minnesota, Pennsylvania, Alaska, Arizona, California, Maryland, Illinois, and Florida.   

    On the third point, the additional contributions to the pension system are resulting in long-term freed-up spending because of lower annual pensions contributions.

    These savings, currently $730 million per year, allow our state to spend predictably toward social and economic programs over the next 20 years in a way that would not be possible without the pension contributions.

    While future use of surplus volatile income is subject to debate, it should be noted there are tangible benefits that multiple generations of Connecticut residents will experience because of the current system. 

    Evaluating Alternative Volatility Cap Methods 

    In response to continued surplus volatile revenue, multiple reports have been made suggesting possible “tweaks” to the volatility cap annual adjustment that should, in theory, result in more revenue available to lawmakers on a year-to-year basis.

    In the following section, we evaluate the approach for a dynamic cap offered by Yale’s Tobin Center.

    This evaluation should not be considered a criticism, but rather a perspective on some of the possible advantages, drawbacks, and practical considerations that legislators should keep in mind when considering alterations to the existing cap. 

    Dynamic Cap 

    The existing cap structure was created in 2017 to set the initial cap of $3.15 billion and amended in 2018 to allow for the adjustment of the cap based on the five-year compound annual growth rate in personal income. 

    A common criticism of this structure is that it was set at a single point in time, has not adjusted to changing revenues, and perhaps should be adjusted to accommodate a “new normal” in volatile revenues.

    One such recommended adjustment that has gained attention is the “Dynamic Cap” method proposed by the Yale Tobin Center.

    Under this approach, previous years’ volatile revenues would be adjusted to current price levels and averaged to create either a five-year or 10-year rolling average.

    In theory, such a method should result in a more responsive cap that frees up more revenue while still allowing for more modest surpluses to be put toward our Budget Reserve Fund and pension systems. 

    The chart below recreates their method. 

    One important thing to point out is that this chart only represents the difference between the current cap and the proposed cap.

    In two of the fiscal years in question, revenues would have fallen below the cap, thus some of the cap would not have been “freed up”. 

    For example, in 2023 a five-year dynamic cap created under this methodology would have been set at $5.484 billion, but total volatile revenues in fiscal 2023 were only $4.954 billion. 

    The following chart illustrates the true freed up revenue based on available revenues. 

    The purpose of highlighting this nuance is to acknowledge that the cap is not destiny.

    In years where the cap is set above where actual revenues come in, the budget is limited to those revenues.  This is true regardless of which cap structure is agreed upon.  

    Another important consideration when evaluating differences between cap methods is that the price deflator being used to determine the cap can drastically impact the cap in any given year.

    The Tobin Center cap adjusts for income by utilizing personal income data from the Bureau of Economic Analysis to create an index.

    In the following charts, we illustrate how much freed up revenue a rolling cap would create using CPI or no price adjustment as well.   

    While the differences between different price adjustments might be subtle, over this six-year period the price adjustment being used can alter spending or saving by up to $1.5 billion.  

    Pension Contributions Impact 

    The debate around how to spend excess volatile revenues has been characterized as a dilemma of spending on social programs versus saving toward pensions.

    This mischaracterizes slightly how pension contributions ultimately impact the state budget and warrants an expanded analysis.   

    Due to the significant unfunded liabilities in the state’s pension systems, lawmakers came to an agreement in 2017 to provide additional funds toward these systems to ensure their solvency in the long-term.

    As a result, part of our state budget each year is allotted for Actuarily Determined Contributions, which supplement the funds the state puts into the pension systems based on existing agreements and staffing.

    What this arrangement allows is greater transparency into how the state intends to ensure the solvency of pensions and provides a predictable, stable figure for those contributions over the next 20 years at which point the unfunded liability would be eliminated.

    Much like any other debt with a fixed-payment schedule, additional contributions today reduce the size of future required payments.

    As a result of the contributions to the pension system, Connecticut will save $737 million per year for the next 20 years.

    OPM estimates that for each $100 million contributed to the funds, $8.5 million is saved in ADCs each year over a 20-year span, freeing up the budget for other expenditures.

    As a result of the contributions to the pension system due to our current volatility cap structure, Connecticut will save $737 million per year between fiscal 2024 and fiscal 2045. 

    What this illustrates is that making additional pension contribution is not like placing money into a savings or investment account in the traditional sense.

    In effect, these contributions increase stable spending on social programs with interest over a 20-year period.   

    As such, when we are evaluating different volatility cap structures with the expressed intent of maximizing social spending, it is crucial to consider these annual savings in those evaluations.

    Given that changes to the cap that would increase spending in the near term come at the expense of savings, we can estimate how much additional pension contributions different cap structures would create, compare those to the savings under the current system, and create a “Freed Up Spending” estimate. 

    The resulting figure highlights the importance of considering these pension savings in our calculations. 

    We estimate that under the five-year and 10-year dynamic caps between 2018-2023, additional pension contributions would have resulted in $1.684 billion and $3.085 billion and annual savings toward ADCs of $143 million and $258 million respectively.

    That compares to the current cap system that deposited $7.6 billion and resulted in annual savings of $651 million.

    We can use the difference between the savings under the proposed caps and the current cap to highlight the true difference in available spending.

    Putting more money toward pensions is not denying those funds to social programs. 

    Prior to adjustment, the five-year rolling average would result in significantly more available funding in 2023.

    However, as the figure illustrates, the additional pension savings from the more conservative 10-year rolling average ultimately overcome the cap differences to allow for greater spending.

    This highlights again that the volatility cap is not destiny.

    Lawmakers are ultimately constrained both by the actual revenues they receive and existing budget obligations.

    Putting more money toward pensions is not denying those funds to social programs. Rather it is allocating those funds to social programs in the future. 

    Present vs. Future Spending 

    As we illustrated in this section, pension contributions represent the act of allocating spending to future years rather than the present year.

    So how should we evaluate different cap systems with that in mind with the hope of maximizing functional spending?

    This is a difficult question considering the volatile nature of EF and PTE revenues and the unique market conditions that have influenced those revenues since fiscal 2018.

    However, we can game out some possible scenarios over the next several years using OPM revenue projections to help us. 

    Under the first scenario, let’s presume that OPM revenue estimates are correct as are their estimates for personal income.

    The figure below illustrates the projected revenues compared to the different caps. 

    As we can see, both the five-year rolling average and 10-year rolling average caps would be greater than projected revenue over this span.

    In 2028, the five-year rolling average would decline as it removes the high-water mark of 2022 but stays above the projected revenue.

    In this scenario, the dynamic caps allow an average of $1.3 billion in additional spending per year and produce zero surpluses while the current cap structure produces $1.3 billion annually in surpluses. 

    Assuming all projected general fund surpluses go toward topping up the BRF, thus leaving the volatility adjustment to go toward pensions, the current cap system would deposit $5.23 billion to the pensions and the dynamic caps would contribute none.

    Incorporating the freed-up spending both from previous pension contributions and these new contributions, the following chart shows actual available spending under the three caps. 

     As we can see, while true spending remains higher than the current cap structure, it declines each year due to the declining ADCs created by additional pension contributions.

    It also illustrates that there is a theoretical breakeven point where the reduced ADCs under the current system would be greater than the freed-up revenue under the proposed caps.   

    Let’s consider a scenario where there is a 20% decline in revenue in year one followed by a slow return to baseline revenue.

    This would mirror a similar scenario, albeit slightly more optimistic, compared to the early 2000’s internet crash.   

    Under this scenario, overall spending under the dynamic caps is lower during the first year of the shock before recovering.

    Spending relative to the current cap structure actually exceeds the business-as-usual scenario due to reduced pension contributions from lower revenues.

    Additionally, in each of the years of the shock scenario, the dynamic caps remained higher than expected revenues. This illustrates that dynamic caps adjusted for prices can result in a cap that is structurally higher than projected revenue over multiple years and may not effectively respond to changing market conditions.

    Under this scenario, if legislators were budgeting based on expected revenue under the cap, the shock would have likely resulted in a $2 billion shortfall over these four fiscal years using dynamic caps and no shortfall under the current cap. 

    Conclusion and Recommendations 

    Volatile revenues in Connecticut’s budget process remain difficult to address.

    These revenues by their very nature create uncertainty that must be accounted for when responsibly planning future spending.

    Much like the financial industry from which these revenues are derived, the warning that “past results do not indicate future performance” should remain the motto when analyzing the effectiveness of our revenue controls.

    As such, any consideration of adjustments to the volatility cap should keep several points in mind: 

    1. These revenues remain volatile and can fluctuate up to 20% in any given year. 
    2. Revenues received in the past several years have been impacted by a unique market and political environment and it remains to be seen the extent to which these conditions will continue. 
    3. The volatility cap is not a forecast.  Revenues available for spending are still subject to what revenues arrive in the state’s coffers. 
    4. Pension contributions have a significant impact on future spending and any cap adjustments should consider the impact on required pension ADCs. 

    Given the uncertainty of future revenues, alternative approaches to handling these revenues could include: 

    1. Earmarking future savings in ADCs for specific purposes, thereby securing funding for certain social programs. 
    2. Broadening the available uses for surplus funds, but requiring those funds only be budgeted after the close of the fiscal year once revenues have been received. 
    3. If a dynamic cap is to be used, only use the nominal figures from previous years without a price adjustment to avoid setting a cap arbitrarily high. 
    4. Increase the bond caps relative to the size of the pension contribution that year.  i.e. if $1 billion is deposited then $500 million in additional bonding is available thereby reducing the long-term debt of the state while enabling near-term spending. 

    For more information, contact CBIA Foundation for Economic Growth & Opportunity director Dustin Nord (860.244.8522).

    Tags:

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    Stay Connected with CBIA News Digests

    The latest news and information delivered directly to your inbox.

    CBIA IS FIGHTING TO MAKE CONNECTICUT A TOP STATE FOR BUSINESS, JOBS, AND ECONOMIC GROWTH. A BETTER BUSINESS CLIMATE MEANS A BRIGHTER FUTURE FOR EVERYONE.