On Jan. 11, 2024, S&P Global Ratings published a request for comment (RFC) on its proposed revisions to the approach it uses to rate U.S. governments. Following feedback from the market, we finalized and published our criteria, titled "Methodology For Rating U.S. Governments," on Sept. 9, 2024.

We'd like to thank investors, issuers, and other intermediaries who provided feedback. This RFC process summary provides an overview of the changes between the RFC and the final criteria as well as the rationale behind those changes.

External Written Comments Received From Market Participants That Led To Significant Analytical Changes To The Final Criteria

Comment:  One commenter recommended that the methodology reflect whether state or local governments assess the need for capital investments and measure and report liabilities for current and future deferred maintenance and preservation needs. They suggested this could be assessed as part of the Management: Long-term planning subfactor or as an adjustment to the Management factor. The commenter also recommended we assess a government's pension stress testing and risk reporting as an adjustment within the Debt and liabilities, Management, or Institutional framework sections of the criteria.

Response:  While the proposed criteria would have provided the flexibility to reflect the credit support provided by exceptionally strong reporting (for example, in the anchor decision or the holistic analysis), this comment brought to our attention a circumstance for improving the management assessment that could be somewhat common. This led us to update our transparency and reporting adjustment within the Management factor such that it can both improve and worsen the initial assessment. We also added an example to the adjustment that says, "Exceptionally strong reporting on areas outside of standard financial reports that management uses to meaningfully reduce financial risk." This new example could be used to reflect the circumstances mentioned by the commenter and improve the final management assessment.

Comment:  In another comment, the same commenter noted that states with greater revenue volatility need to save more than those with less fluctuation. They asked that we consider budget stress tests to set reserve targets within our assessment of reserves and liquidity.

Response:  We view the use of a budget stress test to set reserve targets favorably and a sign that management has well-defined policies that reflect the operating environment of the government. We reflect such governance attributes in our assessment of management within the Policies subfactor. We can also assess the volatility of the revenue structure within our assessment of financial performance. For our Reserves and liquidity factor, we believe it's appropriate to set fixed thresholds, where higher reserves receive stronger assessments and lower reserves, weaker. If there is evidence that a state's reserve targets don't match the volatility or cyclicality of the operating environment, analysts may qualitatively adjust their assessment of reserves and liquidity. In addition, having a mechanism that replenishes reserves, which is frequently part of a state's budget stress test, is also an important part of our assessment of reserves and liquidity.

While we didn't make a change directly in response to this comment, this comment did lead us to clarify our description of the initial assessment of reserves and liquidity for states indicating that the 8% reserve is an annual target. This change only reflects a clarification of our analysis and isn't a change in the approach in the RFC.

External Written Comments Received From Market Participants That Did Not Lead To Significant Analytical Changes To The Final Criteria

Scope and overall framework

Comment:  One commenter noted that the methodology 1) provides a transparent approach to evaluating all U.S. governments and 2) is organizing and making more transparent credit considerations already embedded in the rating process, as the low level of expected rating changes demonstrates. They believe this alignment is helpful for governments and other market participants to analyze S&P Global Ratings' rating rationales.

The commenter also noted that the proposed methodology should help K-12 school leadership teams better understand how their school compares to other governments, adding that the enhanced transparency would allow all issuers to better understand credit pressures and the credit factors in our criteria that are under the leadership teams' control.

Comment:  Another commenter asked that we provide more guidance on and examples of the types of special districts under the scope of these criteria.

Response:  In our definition of local governments in the glossary of the criteria, we list examples of various special districts that could fall in the scope of this criteria. These include library, park, fire, and forest preserve districts. Governments such as water districts would fall in the scope of the U.S. municipal water, sewer, and solid waste utilities criteria, not the U.S. government methodology. Local governments and enterprise entities have distinct business and financial profiles that we evaluate in different criteria. The nature of operations will be a key determinant of which criteria we apply.

Modifiers

Comment:  Three commenters had concerns about how governments that serve a small population would be treated under the criteria, noting the potential for undue penalization due to size.

Response:  We've observed meaningful differences in the rating performance of smaller governments versus larger ones. When in credit distress, it's our view that multiple credit factors pertaining to governments with small populations tend to deteriorate more quickly, often leading to larger rating transitions. Because the individual credit profile (ICP) factors are largely agnostic on the scale of operations, and to ensure we're taking a forward-looking view of creditworthiness, we believe it's appropriate for our methodology to enable us to capture the vulnerabilities of smaller governments. Accordingly, we typically lower the anchor rating on smaller governments by one notch. This allows us to differentiate smaller governments from larger ones with a similar ICP. Importantly, the inclusion of the modifier doesn't meaningfully alter the distribution of ratings.

Individual credit profile--economy

Comment:  We received feedback from two commenters that the local economic profile adjustment doesn't go far enough to reflect cost-of-living differences across the U.S. states and metropolitan areas. Instead, they suggested that it would be helpful to use the U.S. Bureau of Economic Analysis' Regional Price Parities (RPP).

Response:  We agree that using RPP captures the impact of differences in the prevailing price levels across the U.S. and can add value to an assessment of economic activity. However, we're not altering our approach as presented in the RFC. We use inflation-adjusted measures of gross product in the initial assessment to capture regional inflation effects. Moreover, in the local economic profile adjustment of the criteria, we assess a local government's income relative to county income. We believe this approach minimizes the effects of regional price differences because it compares a local government's income to its county average.

In addition, governments earn revenues based on nominal incomes or nominal prices of goods and services consumed. Then, they use these revenues to service financial obligations in nominal terms. Therefore, we believe it's appropriate that our initial assessment considers nominal per capita incomes relative to the U.S. In this case, from our perspective, the application of RPP wouldn't provide any additional insight in assessing the government's creditworthiness.

Comment:  One commenter stated that reducing the economy factor weight to 20% from 30% creates an implicit bias against governments with significantly large economies. This commenter encouraged the criteria to be sufficiently flexible to incorporate into the rating outcome the unique credit strengths of such governments.

Response:  The 20% economy subfactor weights align with the methodologies we use to assign ratings to U.S. states and large local and regional governments outside the U.S. We believe the methodology provides analytical flexibility in other parts of the analysis to ensure we can appropriately reflect each government's unique credit characteristics.

Individual credit profile--reserves and liquidity

Comment:  We received market feedback supporting the separation of reserves and liquidity into their own ICP factor rather than treating them as part of a broader budgetary performance factor, as is the case in our U.S. state ratings methodology. The commenter noted that reserves are essential for managing downturns and maintaining structural balance over the long term.

Comment:  One commenter asked us to consider an additional adjustment in the reserves and liquidity factor to assess the stringency of reserve-replenishment provisions. They noted that in some cases, statutes require states to pay back reserves in a short amount of time, making them inaccessible for a deep, multi-year recession.

Response:  We believe the criteria provide sufficient flexibility to handle governments' unique credit characteristics and nuances. States have various tools to address a prolonged economic or fiscal stress scenario. We believe the criteria as proposed allow us to consider if the stringency of reserve replenishment would have any negative implications for financial performance, reserves and liquidity, and management.

Individual credit profile--management

Comment:  We received market feedback expressing support for the inclusion of "preparedness for acute or chronic evolving risks, such as extreme weather, natural disasters, or cyber security events," as factors that can either improve or worsen creditworthiness.

Comment:  One market participant suggested that we provide additional detail about how we evaluate preparedness for extreme weather and natural disaster risk, with more explicit guidance about what type of related actions we view as a positive credit factor.

Response:  We believe this methodology strikes the right balance between identifying the relevant credit factor while providing flexibility to consider each government's unique credit characteristics. Our evaluation of preparedness for acute or chronic evolving risks--such as extreme weather, natural disasters, or cyber security events--is in the context of the government's operating environment. We also consider how a government compares to peers. We expect that experienced management teams will adapt and change their strategies as these risks evolve. And because they are evolving, we will reflect these risks or opportunities in our ratings only when we have sufficient visibility and certainty that an entity's preparedness--or lack thereof--could materially influence its creditworthiness.

Comment:  One commenter asked that in our evaluation of management, if we believe a government faces a challenging management and governance environment, has an understaffed management team, or lacks relevant skills or experience, we consider third-party support for certain technical functions.

Response:  We agree that third-party assistance can provide meaningful support to government management teams. However, to the degree that state or federal assistance is formalized, we generally consider the role of this support within the revenue/expenditure balance and system support subfactor in our Institutional Framework analysis. If the management team is meaningfully using other technical assistance available to support its practices or policy development, we will account for that within our overall assessment of management.

Individual credit profile--debt and liabilities

Comment:  One commenter recommended updating the debt and liabilities assessment to directly incorporate a metric for pension contribution adequacy and sustainability.

Response:  We believe that our assessment of pension contribution adequacy and sustainability will be captured in the adjustment for under- or overstated current costs, which includes the example: "Actual pension payments significantly above or below actuarial recommendation or minimum funding progress threshold while considering our discount rate guideline and pension contribution methods such as amortization period, length, and basis (improve or worsen)." This allows for the adjustment of the debt and liabilities assessment to reflect our view of pension contribution adequacy.

Comment:  One commenter recommended updating the debt and liabilities assessment to incorporate an initial assessment or an adjustment for whether pension or other post-employment benefit (OPEB) contributions are set actuarially.

Response:  We agree that setting contributions on an actuarial basis is best practice. We believe the criteria allow analysts to worsen the debt and liabilities assessment as appropriate when pension payments aren't in line with actuarial recommendations or our minimum funding progress metric. Assessing the size of the contribution relative to budget and then adjusting our assessment to reflect how this payment compares to the actuarially determined contribution or our minimum funding progress metric allows us to better rank-order this factor's relative risk.

Comment:  Two commenters asked why we use debt per capita rather than debt relative to budget or revenue generating capacity to measure a government's debt burden.

Response:  The criteria does account for debt to budget with the metric "Current cost for debt service and liabilities." This measures the annual debt service cost relative to total governmental funds revenue and is 50% of the initial debt and liabilities assessment. In addition, the criteria consider long-term debt liabilities relative to budget through the qualitative adjustment for under- or overstated long-term debt and liabilities. This is to account for instances when per capita metrics overstate the government's debt burden. We chose debt per capita as a complementary metric because it's widely available, straightforward, and allows the criteria to differentiate among governments.

Comment:  Two commenters requested that we remove tax increment financing (TIF) and special assessment revenue bonds from our calculations of an issuer's direct debt if the issuer has no exposure when there's a revenue shortfall.

Response:  It's long been our practice to include TIF and special assessment bonds as direct debt of the issuer. While these bonds are typically passively managed and payable from dedicated revenues, they're obligations undertaken by the government and widely used for economic development. The projects associated with these instruments are typically infrastructure benefitting the governments and payable from taxes. Therefore, we maintain our view that these are direct financial obligations, and we don't offset them when calculating the net direct debt burden.

Comment:  One market participant questioned whether some or all of the qualitative adjustments within the debt and liabilities factor are subject to half-point adjustments.

Response:  As stated in the third paragraph of the Individual Credit Profile section, all of the qualitative adjustments within this factor are subject to half-point adjustments. We chose to incorporate half-point adjustments because 50% of the factor is current costs, and the other 50% is long-term liabilities. By using half-point adjustments, our analysis can more clearly take into account which adjustments are related to current costs as opposed to long-term liabilities.

Other Substantive Market Feedback That Led To Significant Analytical Changes To The Final Criteria

Comment:  One commenter asked us to include our pension funding guidelines in the criteria.

Response:  In the criteria appendix--within the definitions for amortization basis, amortization length, and amortization period--we provide our views on pension and OPEB contribution methods. We state that we generally view the following as weak: an amortization period that assumes growth higher than inflation, amortization payments defined over periods longer than 20 years, and open amortization methods.

However, to increase the visibility of these definitions, in the final criteria we relocated these three definitions into the definition of "Pension and OPEB contribution methods."

Other Substantive Market Feedback That Did Not Lead To Significant Analytical Changes To The Final Criteria

Comment:  One commenter questioned whether the criteria would include a rating cap when a government exhibits a structural imbalance.

Response:  The criteria no longer cap a rating due to a structural imbalance alone. Rather, through the financial performance and management factors of the ICP, the criteria provide the analytical flexibility to account for a government with a structural imbalance. If the management assessment is '5' or worse, we can lower the anchor rating to reflect the distress. If the management assessment is a '6', we cap the rating at 'BBB+'.

Individual credit profile--debt and liabilities

Comment:  One market participant asked how the criteria would handle the issuance of pension obligation bonds (POBs), specifically if the bonds will fund 100% of the unfunded liability.

Response:  Within the initial assessment for debt and liabilities, we treat unfunded pension liabilities and debt obligations equally. Therefore, if a government were to issue POBs in an amount that would fully fund its pension obligation, the government's initial assessment for pension liabilities per capita would improve, while its initial assessment for debt per capita would worsen equally. The initial assessment for current costs includes both debt service and pension contributions. Therefore, after accounting for one-time payments, the annual increase in the debt service payment would be offset by the annual decrease in pension contributions. If there were other changes to pension plan assumptions or costs, we would reflect them through qualitative adjustments to the initial assessment, such as "under or overstated current costs" or "projections that suggest a different assessment."

Comment:  One market participant suggested that within our adjustment for "Under or overstated long-term debt and liabilities," we update our example that says "Understated population, leading to overstated liabilities per capita, such as a significant secondary home market (improve)" to include entities with a significant commercial or industrial base.

Response:  A rating committee's determination that debt is overstated for governments with significant commercial or industrial bases will be context dependent. While this is a possible adjustment, we don't feel it will be sufficiently common to warrant inclusion as an explicit example. We intend for the adjustment examples to reflect our typical adjustments, not cover all cases.

Comment:  One commenter suggested that in our initial assessment of current cost within the Debt and Liabilities factor, we use the interest on a government's net pension liability (NPL) and net OPEB liability (NOL) instead of the annual pension and OPEB contributions. This would reflect the annual cost of maintaining the NPL or NOL.

Response:  Our approach to pension and OPEB liability payments looks at current scheduled payments because that's the clearest expectation of future cash flows that could affect the budget. However, for pensions, we compare those actual payments with a calculation that uses interest on a government's NPL as well as an additional amount representing meaningful progress in the repayment of those liabilities (our minimum funding progress--MFP--metric). We do this to assess the sufficiency of current pension costs. And if there's information on OPEB, we use it to inform our opinion of OPEB cost sufficiency as well. If a government were to pay only interest and no principal in a given year, we would qualitatively worsen our assessment because current costs are understated. Similarly, we may also qualitatively worsen our assessment if a government's actual pension contributions are lower than its actuarially determined contribution or our MFP metric. We assess the size of these payments relative to budget to gauge the effect these costs have on a government's overall creditworthiness.

Comment:  One market participant questioned how we will assess the risks a government takes in its pension fund and how that might affect the future affordability of its NPL.

Response:  Within the debt and liabilities factor, we can adjust for under- or over-stated long-term debt and liabilities. We may worsen the debt and liabilities assessment for governments with elevated NPL per capita and high pension discount rates that could increase future liabilities or lead to significant cost volatility from amortization costs. We limited this adjustment to governments with elevated NPL because we found that applying it more broadly over-penalized governments with strong funded ratios and very limited liabilities. Further, if we believe that pension assets are built around opaque risks we see increasing within general pension asset portfolios in the U.S., we could also worsen the assessment using the adjustment for projections that suggest a different assessment, even if the discount rate is low.

Significant Analytical Changes To The Final Criteria That Did Not Arise From Market Feedback

We made analytical changes to the final criteria that didn't arise from market feedback. However, we don't consider these changes to be material to the analytical framework.

Modifiers

We added the word "generally" to the modifier for local governments with effective buying income greater than 150%. This was to add more discretion around when this modifier is applied to governments near the threshold. We believe this will improve rating stability and allows for a more forward-looking application of the modifier.

Individual credit profile--economy

We clarified which initial assessment metrics use inflation-adjusted (real) data versus nominal values that don't incorporate such an adjustment. This clarification was only to provide additional transparency about the metrics we use for analyzing government economies and doesn't reflect any change to the metrics in the RFC.

Individual credit profile--Debt and liabilities

We clarified the first example under the adjustment for under- or overstated liabilities that can improve the initial assessment for governments with tax-secured debt that have demonstrated sufficient self-support from utilities. It now specifies that this adjustment only applies to local governments and not states. This change was only to clarify our analysis and isn't a change to the approach in the RFC.

Appendix 1: Debt Statement Analysis

We made two changes to Appendix 1:

• In accordance with recent changes to accounting standards, we added language clarifying that we include subscription-based IT arrangements in our calculation of net direct debt.
• We introduced the term "offsetting debt" to distinguish "offsetting debt" from "self-supporting debt." By definition, offsetting debt is subtracted from the government's gross direct debt to calculate its net direct debt. It includes debt such as enterprise debt secured by revenues only, moral obligation debt that hasn't required any contribution to the debt service reserve fund, and tax anticipation notes. Self-supporting debt is tax-secured debt supported by an enterprise such as water, sewer, solid waste, and electric utilities, where such support is consistently sufficient. Self-supporting debt is included in the government's net direct debt, but the initial debt and liabilities assessment could be improved through a qualitative adjustment to reflect the utility support. In the RFC, we'd referred to both types of debt as "self-supporting."