Introduction

The Workforce Innovation and Opportunity Act of 2014 (WIOA; P.L. 113-128) is the primary federal workforce development statute.¹ Title I of WIOA authorizes three formula grant programs that support state workforce systems.² State workforce agencies use these funds to meet local workforce needs, including providing career services and training to individual workers.

WIOA authorized appropriations for FY2015 through FY2020. Since FY2021, the systems and programs authorized by WIOA have continued to be funded through the annual appropriations process. Congress has engaged in ongoing efforts to reauthorize WIOA, including a House-passed reauthorization bills in each of the 117^th and 118^th Congresses, but no consensus has emerged.³

Title I of WIOA authorizes three formula grants—Adult Activities, Dislocated Worker Activities, and Youth Activities—that target specific populations. The Adult Activities grant is the most flexible of the three formula grants. States can use these funds to support any worker age 18 or over, though certain statutory provisions give priority for more intensive services to certain low-income individuals and other disadvantaged subpopulations.

The formula that allots Adult Activities funds is complex and its nuances may not be well understood. During the WIOA reauthorization process, numerous questions have arisen related to the design of the Adult Activities formula and how it allocates funds. These include questions about how the specialized formula factors are derived and how the existing factors compare to more traditional indicators. Other questions have arisen about alignment between the formula factors and the populations the program serves and the sources of year-to-year volatility in states' formula funding.

In support of WIOA reauthorization, this report uses a combination of qualitative and quantitative analyses that are designed to address common inquiries and provide policymakers with a comprehensive understanding of how the Adult Activities formula is designed, how it is implemented, and how it allocates funds to state workforce agencies.⁴

The report begins with background on the WIOA statute to provide context, situate the analysis within the federal workforce strategy, and provide a brief legislative history.⁵ It then provides a brief description of the formula to establish terminology and key principles.

The bulk of the report presents a series of qualitative and quantitative analyses with the goal of illuminating each step of the formula, from the initial development of the formula factors to the final application of the adjustment provisions. The analysis is divided into five sections. Each section has somewhat discrete methodologies and findings. The introduction to each of the five sections summarizes the major findings associated with the accompanying analysis.

The report concludes with a discussion of potential options for revising the formula. Options to revise the formula factors include adopting factors that are more transparent or more closely aligned with the program population. Each of these factor strategies options has advantages and potential drawbacks. Some potential factor changes may have a large distributional effect. The final section of the report discusses policy options to limit the immediate effects of any factor changes on states' grants.

Context and Background

WIOA authorizes grant programs and state systems that coordinate federal funding streams to meet local labor market needs. Title I of WIOA authorizes formula grants to states that support the operation of state workforce systems. These funds are allotted to states based on their relative shares of specified formula factors. States can use the funds to operate a system of One-Stop Career Centers that provide career services and training. States and One-Stop Career Centers coordinate these activities with other federal funding.⁶

Title I of WIOA authorizes three formula grant programs that support workforce development interventions for specified populations. Youth Activities supports training and other workforce preparation services for in-school and out-of-school youth who are low-income or meet other criteria. Dislocated Worker Activities supports reemployment efforts for workers who have been terminated or laid off and meet other criteria. Adult Activities funding (the focus of this report) has the most flexibility in the populations it serves. It can be used to provide career services to any person age 18 or over, though it prioritizes low-income workers and other disadvantaged populations. To receive training under the Adult Activities program, a participant must demonstrate a need for training to obtain or retain employment that leads to self-sufficiency.⁷

WIOA formula funds are appropriated on a fiscal year basis but support activities that operate on a program year (PY) basis. The program year runs nine months behind the fiscal year (e.g., PY2020 ran from July 1, 2020, through June 30, 2021). Funds appropriated in a given fiscal year support activities in the corresponding program year (e.g., funding from FY2021, beginning October 1, 2020, supports the operation of programs in PY2021, beginning July 1, 2021).⁸ Post-appropriation activities (such as formula allotments) are typically reported on a PY basis.⁹ Most of this report will discuss the allocation of funds associated with program years.

Table 1 presents annual funding for WIOA Title I formula grants since the law's enactment. In PY2023, funding for Adult Activities formula grants to states was about $881 million, accounting for roughly 30% of the formula grant funding in Title I of WIOA. The funding level for formula grants in Table 1 is slightly lower than the total appropriation due to set-asides and reservations, including dedicated funding for outlying areas.¹⁰

Brief Legislative History

WIOA was enacted in 2014 and took effect beginning in PY2015. The structure and funding formula for the Adult Activities program in WIOA generally follow the formulas for similar programs in the two major predecessor statutes: the Workforce Investment Act of 1998 (WIA; P.L. 105-220) and the Job Training Partnership Act of 1982 (JTPA; P.L. 97-300). Both WIA and JTPA authorized a formula grant to states that had three factors and similar calculation processes to those used in the current WIOA Adult Activities formula.

A more detailed legislative history and descriptions of formulas in each of JTPA, WIA, and WIOA are in Appendix A.

Brief Overview of the Adult Activities Funding Formula

Funding for the Adult Activities program is allotted to states via formula. The formula allots funds based on states' relative shares of three equally weighted formula factors and then applies several adjustment provisions.¹¹ States may use the funds to support career services and training for program participants. The Adult Activities program is a universal access program (i.e., it has no eligibility requirements), but it has several mechanisms to target services to workers who are low-income and those with barriers to employment.

Formula Factors and Calculation of Initial Grants

Initial grant calculations are made on the basis of each state's relative share of three factors. One-third of the funding for grants is allotted based on each state's relative share of each factor (see the text box for more detail on the concept of relative share). The factors are the following:

• Disadvantaged adults (DA factor). This factor considers the number of adults aged 22 to 72 "who received an income or [are] a member of a family that received a total family income" that does not exceed the higher of the poverty line or other specified thresholds.¹²
• Unemployment in Areas of Substantial Unemployment (ASU factor). This factor considers the number of unemployed individuals in areas of the state that (1) are "of sufficient size and scope to sustain a program of workforce investment activities" and (2) have an unemployment rate of at least 6.5%. In cases where a state's overall unemployment rate is at least 6.5%, the state is considered a whole-state ASU and all unemployed persons in the state are counted in the calculation of the ASU factor. Note that this report will use the term ASU when referring to a geographic area of substantial unemployment and the term ASU factor when referring to the count of unemployed individuals within an ASU that is considered by the formula.
• Excess unemployment (EU factor). This factor considers the higher of (1) unemployed individuals in excess of 4.5% of the civilian labor force in the state or (2) unemployed individuals in excess of 4.5% of the civilian labor force in the ASUs in the state. For a state that qualifies as a whole-state ASU, these two numbers will be identical.

Notably, all of the factors are composite indicators that are developed solely for the purpose of application in the WIOA formulas. The factors cannot be easily constructed or replicated using regularly published data.

Adjustment Provisions

WIOA has several provisions that establish minimums for small states and limit year-to-year fluctuations in individual states' Adult Activities grants. These adjustment provisions are applied after the calculation of each state's initial grant based on the state's relative share of the three formula factors.

• Small state minimum: Each state's funding must equal at least 0.25% of the total funds for grants. If a state's initial grant is less than 0.25% of the total funding for grants, the state's grant is increased to 0.25% of the total amount and other states' grants are ratably reduced.¹³
• Hold harmless: Each state's relative share of funding may be no less than 90% of its relative share of the funding from the prior year. If a state's initial grant is a relative share of total funding that is less than 90% of its relative share from the prior year, the state's relative share is increased to 90% of its relative share from the prior year and other states' grants are ratably reduced.
• Maximum grant/stop gain: No state may receive a relative share of funding that is more than 130% of its relative share of funding from the prior year. If a state's initial grant is more than 130% of its relative share from the prior year, the state's relative share is reduced to 130% of its relative share from the prior year and other states' grants are ratably increased.

The hold harmless and maximum grant provisions are based on relative share, not actual grant levels. This means that if total funding for grants increases from one year to the next (a common scenario), the hold harmless dollar amount is more than 90% of the prior year's grant dollar amount and the maximum grant is more than 130% of the prior year's grant dollar amount.¹⁴

Figure 1. Overview of the Adult Activities Formula

Source: CRS analysis of WIOA Section 132(b).

Overview of Analysis, Recurring Themes, and Issues for Consideration

The following sections discuss the results of an in-depth analysis of the Adult Activities formula and issues for consideration as Congress pursues reauthorization of WIOA. The combination of analyses are designed to provide policymakers with a comprehensive understanding of how the Adult Activities formula is designed, how it is implemented, and how it allocates funds.

The Adult Activities formula factors are unique to WIOA. A recurring theme in the analysis is how the statute-specific formula factors compare to more traditional labor market indicators. An implicit question underlying these comparisons concerns what advantages or disadvantages the complex and somewhat opaque indicators established in WIOA offer over more common and perhaps more widely understood indicators.

The report includes both qualitative and quantitative analysis. The qualitative analysis is divided into two parts. The first part reviews the implementation of the formula. It focuses on the complex procedures necessary to develop formula factors that are in compliance with the statutory requirements, including some practices that may not be widely understood. The second part compares the populations that are reflected by the formula factors with the populations that the Adult Activities program serves and prioritizes.

The quantitative analysis is divided into three related parts. The first part analyzes the formula factors relative to more traditional labor market indicators. It aims to identify the characteristics of states that benefit more and less from the complex statutory factors in the Adult Activities formula compared to more traditional indicators. The second part calculates states' year-to-year changes in their relative shares of the formula factors. The third part analyzes the role of the adjustment provisions (e.g., hold harmless) in determining final grant levels and measures the effects that these provisions have on final grant levels.

Formula Factor Design and Implementation: Compliance with Statutory Provisions Requires a Complex Process

This section describes the statutory requirements of each of the three formula factors and how these requirements are implemented. As noted previously, the Adult Activities formula factors are not regularly published metrics but rather are developed exclusively for use in the formula. One of the three factors (the DA factor) is a population-based factor. Two of the factors (the ASU factor and EU factor) are based on specialized measures of unemployment.

This section synthesizes published and unpublished sources to offer a thorough overview of how the three formula factors are developed. Developing these factors in accordance with their statutory specifications requires federal agencies to engage in a series of calculations, sometimes using nontraditional metrics or delegating certain calculations to state agencies. The complexity of these metrics and their associated development means the formula is less transparent than a formula that uses commonly understood and widely available metrics.

Disadvantaged Adults Factor

The DA factor uses published data on persons in poverty in a specified age range as a baseline, but then introduces complexity primarily by adding additional income thresholds that may increase the factor population.

Definition of DA Factor

Section 132 of WIOA defines a disadvantaged adult as

an adult who received an income, or is a member of a family that received a total family income, that, in relation to family size, does not exceed the higher of—

(aa) the poverty line; or

(bb) 70 percent of the lower living standard income level.¹⁷

Other provisions in Section 132 establish that for the purposes of the factor, an adult is an individual between the ages of 22 and 72.¹⁸ The DA factor definition includes additional provisions that require, in the construction of the factors, "to the maximum extent practicable, exclu[sion of] college students and members of the Armed Forces[.]"¹⁹

Statute does not establish a specific reference period for the factor, but it does specify that "all data relating to disadvantaged adults ... shall be based on the most recent satisfactory data from the Bureau of the Census."²⁰

Implementation of the DA Factor

The Census Bureau provides DOL with updated estimates of disadvantaged adults every five years. Data are based on a special tabulation of American Community Survey (ACS) data that consider a five-year reference period. The data, including some intermediate calculations, are published on the DOL website.²¹ DOL publishes a quinquennial Training and Employment Guidance Letter (TEGL) that focuses on these data.²² It is not clear if the complexity of the calculations contributes to the infrequency of the updates.

Reference Periods and Timeliness

The reference periods for each set of DA factors that have been used since the effective date of WIOA are listed in Table 2. The five-year reference period and the application of data for five years can create a gap between the data that are used for awarding funds and the applicable award year. For example, DA data from the 2011-2015 period were used to allocate Adult Activities funds for each of PY2018 through PY2022.

While the DA factors are updated infrequently, the changes in states' relative shares of this factor tend to be somewhat minor when the factors are updated (see the "Year-to-Year Variation in the Disadvantaged Adults Factor is Limited" section).

Calculating Disadvantaged Adults with Incomes Below 70% of the Lower Living Standard Income Level

A distinct characteristic of the DA factor is the statutory requirement that the factor consider individuals in poverty as well as persons with incomes below 70% of the Lower Living Standard Income Level (LLSIL). The latter metric generally allows the DA factor to capture additional adults with incomes above the poverty line but below the specified LLSIL metrics. The use of the additional LLSIL income metric is likely the greatest source of complexity in the DA factor.

The LLSIL is a series of income thresholds determined by family size and geographic location. There are LLSIL metrics for about 35 regions and metropolitan areas.²³ The differences in the metrics across regions and metro areas generally reflect differences in cost of living between the regions and metro areas.²⁴

DOL last fully updated the LLSIL metrics in 1981, but the department uses the Consumer Price Index (CPI) to update the numerical metrics each year solely for use in the WIOA formulas.²⁵ As such, the LLSIL metrics reflect regional differences in cost of living as of 1981, adjusted for inflation.²⁶ In the most recent publication of the LLSIL metrics that were applied to the calculation of DA factors, many (but not all) geographies had family sizes with LLSIL levels where 70% of the LLSIL was above the poverty line.²⁷

Census applies the LLSIL thresholds to data from the ACS to calculate the number of persons from households with incomes below 70% of the applicable LLSIL. These data are then combined with the poverty data to create a data set of persons with incomes below poverty or below 70% of the applicable LLSIL.²⁸

Other Adjustments to the DA Factor

Subsequent adjustments are made to exclude college students and active-duty military, per statute.²⁹ Other adjustments are made for the three states (Kentucky, Minnesota, and Wisconsin) where separate income metrics apply.³⁰

Analysis of the DA Factor

Because the DA factor is only updated once every five years, each state's relative share of the factor is fixed for five years when the factor is updated. For example, in each of PY2018 through PY2022, Ohio's relative share of the DA factor was 3.413%.³¹ This also means that, after the DA factor is updated, a state can accurately predict the factor that will be used to initially allot one-third of the Adult Activities funding for the remainder of the five-year period.

The combination of the five-year reference period for calculating the DA data and the five-year period for using each set of data means that in the final year that a data set is used, grants will be partially allotted on the basis of data that are more than 10 years old. For example, the most recent DA data are based on data from the 2016-2020 period. These data were first used in PY2023 and will continue to be used through PY2027.³²

Unemployment in Areas of Substantial Unemployment

The ASU factor is an estimate of unemployed individuals within specified geographies. In cases where a state's unemployment rate is at least 6.5%, the state qualifies as a whole-state ASU and the state's ASU factor is simply its total unemployment. In cases where a state's unemployment rate is less than 6.5%, the factor captures only the number of unemployed workers in substate ASUs with an unemployment rate of 6.5% or higher. DOL delegates the responsibility of identifying ASUs to state agencies. Following DOL guidance and using DOL-provided data, the state agencies must construct an area or a series of areas that each have an unemployment rate of at least 6.5%. Trends in ASU data suggest that the states' construction of their ASU factors is largely a strategic exercise to maximize the size of the eligible population rather than identification of traditionally defined regions (e.g., county or city) with high unemployment.

Definition of ASU Factor

The statutory formula allots funds based on the "relative number of unemployed individuals in areas of substantial unemployment in each state." WIOA defines an ASU as "any area that is of sufficient size and scope to sustain a program of workforce investment activities … and that has an average rate of unemployment of at least 6.5 percent for the most recent 12 months, as determined by the Secretary."³³ Statute further specifies that "determinations of areas of substantial unemployment shall be made once each fiscal year."³⁴

Implementation of the ASU Factor

Unemployment in ASUs is not a metric that is otherwise published by federal statistical agencies and it must be constructed annually for its application in the WIOA formulas. Unemployment data used in the ASU calculations are based on the preliminary Local Area Unemployment Statistics (LAUS) data published by the Bureau of Labor Statistics (BLS). The preliminary data are revised in subsequent BLS publications, but the initial preliminary data remain as the foundation for the ASU calculations.³⁵

The ASU factor has been implemented by DOL so that the reference period is the 12 months ending June 30 of the year prior to the program year. (See Table 3.) For example, funds for PY2022 (July 1, 2022, through June 30, 2023) were allotted using the ASU metrics for the period beginning July 1, 2020, and ending June 30, 2021, which corresponds with PY2020.

Whole State ASUs

As noted previously, in cases where a state's unemployment rate is equal to or greater than 6.5%, the whole state is considered an ASU. In these cases, the state's ASU factor equals the total regular unemployment in the state. Qualifying as a whole-state ASU is the only way that a state can have 100% of its unemployment considered in the Adult Activities formula.

Within State ASUs

When a state does not qualify as a whole state ASU, state agencies are responsible for identifying ASUs within the state. BLS provides data and issues annual guidance to state agencies to aid in the construction of the ASUs, but the actual construction is done by states. BLS confirms that state submissions are compliant.

In cases where a state must construct its ASUs, unemployment in portions of the state that do not qualify as ASUs are excluded from the ASU factor. This is effectively a cliff in which 100% of unemployment in constructed ASUs is counted in the ASU factor while 0% of the unemployment in non-ASU areas is counted.

Statute does not constrict ASUs to traditional political boundaries such as cities or counties; it only specifies that an ASU must be "of sufficient size and scope to sustain a program of workforce investment activities."³⁶ DOL guidance has operationalized this provision to require that ASUs have a population of at least 10,000.³⁷

This combination of parameters means that components of ASUs may be (and typically are) as granular as census tracts. This can allow states to strategically construct ASUs in a variety of ways.

In cases where a state does not qualify as a whole-state ASU but the statewide unemployment rate is close to 6.5%, the state can use a subtraction method to create large ASUs or a large single ASU. Under this method, states remove low-unemployment census tracts from consideration until the remaining contiguous tracts yield a collective unemployment rate of at least 6.5% and therefore qualify as an ASU.

Alternately, a state can use an addition method to construct ASUs in the state. This method consists of identifying all tracts with unemployment above 6.5% and adding adjacent tracts with lower unemployment rates until the overall unemployment rate for each ASU averages 6.5%.³⁸

States' strategic construction of ASUs may result in traditionally associated geographies (such as census tracts within a city) being separated. For example, ASUs may be strategically constructed so that lower-unemployment parts of a jurisdiction or area are excluded and higher-unemployment tracts of an adjacent jurisdiction are included, creating a larger area with an unemployment rate of at least 6.5%.

Analysis of the ASU Factor

A critical consideration related to the ASU factor is whether the state qualifies as a whole-state ASU or if the state must construct its ASUs. If a state qualifies as a whole-state ASU, it will have all of its unemployment captured by the ASU factor. If the state's unemployment rate is less than 6.5%, the state must construct its ASUs and some portion of the state's unemployment will fall outside of ASUs and therefore not be captured by the ASU factor.

Since WIOA took effect in PY2015, states have constructed their ASUs in more than 80% of cases. As presented in Table 4, cases where larger numbers of states qualified a whole-state ASUs tend to be concentrated in reference periods when overall unemployment rates are high.

Strategic Construction of ASUs

As noted previously, states that do not qualify as a whole-state ASU must construct ASUs following the procedures described in the "Within State ASUs" section of this report. Optimized construction of these areas typically yields ASUs with overall unemployment rates at or slightly above 6.5%.³⁹ If a state constructed an ASU with an unemployment rate well above 6.5%, it is likely to be to the advantage of the state to add additional census tracts to the ASU, even if those tracts had low unemployment rates that would reduce the unemployment rate of the constructed ASU to closer to 6.5%. As long as the additional tracts contained at least some unemployed individuals and did not reduce the overall unemployment rate within the ASU below 6.5%, the additional tracts would increase the total number of unemployed individuals in the ASU and therefore increase the corresponding ASU formula factor for the state.⁴⁰

Table C-2 presents data on states' unemployment and ASUs for PY2023. Because only one state qualified as a whole state ASU, nearly every state had to construct their ASUs. Several trends emerge from the table, including the following:

• Most states were able to construct ASUs with an unemployment rate of almost exactly 6.5%. Of the 51 states that had to construct ASUs, 46 were able to construct ASUs that had overall unemployment rates that were (or at least rounded to) 6.5%.⁴¹
• Many states constructed a single, contiguous ASU. Of the 51 states that had to construct their ASUs, 27 created a single ASU. This included several large states: of the six states with the largest civilian labor forces, five of them created a single ASU within the state.⁴² These large states' construction of a single contiguous ASU may support the point that the construction of ASUs is a strategic exercise rather than identification of specific areas of economic disadvantage within the state.

Excess Unemployment Factor

The EU factor measures unemployment (i.e., unemployed individuals) in excess of 4.5% of the civilian labor force in either a whole state or within the state's ASUs. In the approximately 80% of cases since PY2015 where a state did not qualify as a whole-state ASU, the state's EU factor was largely a function of the state's ASU factor.

Definition of EU Factor

Statute establishes two definitions of excess number of unemployed and specifies that the higher number applies to the formula allotment. Excess number of unemployed is the higher of

• "the number of unemployed individuals in excess of 4.5 percent of the civilian labor force in the State," or
• "the number of unemployed individuals in excess of 4.5 percent of the civilian labor force in areas of substantial unemployment in such State."⁴³

In cases where a state qualifies as a whole-state ASU per the ASU definition, the two excess unemployment metrics will be the same. The construction of the second factor option (unemployment in excess of 4.5% of the civilian labor force in ASUs) means that every state that has at least some unemployment in ASUs will have a nonzero EU factor because ASUs, by definition, have an unemployment rate of at least 6.5%.⁴⁴

Implementation of the EU Factor

The calculation of each state's EU factor is relatively straightforward. Each state calculates two potential factors based on the two criteria described in statute (see above) and the higher one is used in the allotment formula.

The potential excess unemployment factors are calculated on the basis of (1) statewide unemployment data and (2) data related to the ASU factor. Excess unemployment is calculated by multiplying the size of the labor force in a state (or ASU) by the difference between the actual unemployment rate in the state (or ASU) and 4.5%. For example, if a state has a labor force of 2 million and 7.5% unemployment, the excess unemployment is 3.0% (7.5% - 4.5% = 3.0%) multiplied by 2 million, or 60,000. As noted previously, in whole state ASUs the two EU factors are the same.

The EU factor is updated annually. Like the ASU factor, it has been implemented to consider the 12 months ending June 30 of the year prior to the program year. For example, PY2022 funds were allotted using the excess unemployment metrics for the 12-month period ending June 30, 2021.

Excess unemployment data are published with other factors in annual guidance and on the DOL website. Public-facing data do not include the full breadth of data necessary to recreate the calculation of the EU factor.⁴⁵

Analysis of the EU Factor

As described previously, the EU metric is the higher of two calculations. In the approximately 20% of cases since PY2015 where the state qualified as a whole state ASU, the calculations are identical.

In cases where a state does not qualify as a whole state ASU, the calculations yield differing numbers and the higher metric applies. Analysis of statewide unemployment data and formula factor data shows that, in the vast majority of cases, a state's unemployment in excess of 4.5% of the labor force within ASUs is higher than its statewide unemployment in excess of 4.5% of the statewide labor force.⁴⁶

This trend means that, in practice, the EU factor is typically a subset of the ASU factor. In cases where a state does not qualify as a whole-state ASU, the state's construction of its ASUs becomes particularly relevant because it directly impacts two of the three formula factors that determine a state's initial allotment.

The relationship between a state's relative share of the ASU factor and its relative share of the EU factor varies by overall labor market conditions and how many states qualify as a whole state ASU. The relationship between the EU factor and the ASU factor is discussed in more detail later in this report.

Comparison of Formula Factors and the Adult Activities Population: Alignment is Limited

A potential benefit of complex formula factors may be more precise targeting of federal assistance. This benefit assumes that complex factors are closely aligned with target populations or are otherwise responsive to relevant conditions.

This section compares the populations that are captured by the formula factors with the general population that is eligible for the Adult Activities program and more specific populations that the program prioritizes. In the case of the Adult Activities formula, one of the three factors (the DA factor) demonstrates fairly clear alignment with the program's priority population. The other two factors (the ASU factor and the EU factor) do not specifically target the general population the program can serve or the more specialized population it prioritizes.

Adult Activities: Eligible and Prioritized Populations

The Adult Activities program is somewhat unconventional in that it does not have explicit eligibility requirements. Instead, it is a universal access program (i.e., anyone age 18 or over can access services) that directs states to give priority for more intensive services to workers with certain characteristics ("priority groups"): low-income individuals, recipients of public assistance, and individuals who are basic skills deficient (including English language learners).⁴⁷ The design of the program and subsequent agency guidance have made clear that it should not exclusively serve workers who are members of priority groups.⁴⁸

Review of program data shows that in PY2022, about 67% of total Adult Activities program exiters were members of a priority group and about 60% of total exiters were low-income.⁴⁹ In PY2017, about 55% of total exiters were members of a priority group and about 50% of total exiters were low-income.⁵⁰

One of the stated purposes of WIOA is to support individuals with barriers to employment.⁵¹ WIOA defines this population to include low-income individuals, as well as other populations that may face challenges in the labor market such as ex-offenders, displaced homemakers, older individuals, and long-term unemployed individuals. Neither the Adult Activities program nor other programs authorized under WIOA require states to provide priority of service or any other dedicated priority to workers with barriers to employment who do not qualify as a member of a priority group. For example, while long-term unemployed individuals and displaced homemakers are considered individuals with barriers to employment, they are not considered priority groups under the Adult Activities program.

Disadvantaged Adults Factor

The DA factor, as constructed, is partially aligned with priority groups in the Adult Activities population. It aligns with the low-income priority population but does not consider program participants who are not low-income and does not adjust for the likelihood of labor force participation.

Central elements of the definition of a disadvantaged adult (income below poverty or 70% of LLSIL) align with the low-income criteria for priority groups in the Adult Activities program.⁵² The DA factor does not specifically target any of the other priority group criteria, though it is likely that significant portions of individuals who meet the other priority criteria (such as TANF recipients) are captured by the disadvantaged adults metrics.

The Adult Activities program is designed to support individuals in the labor market, but the DA factor considers individuals in and out of the labor market. This more expansive measure could be seen as increasing alignment with the prospective WIOA population in that it might capture workers who may enter the labor force with the assistance of WIOA services. Alternately, the DA factor, as defined in current law, could be seen as overly inclusive of groups with limited labor force participation. For example, the DA factor considers adults up to age 72. In calendar year 2023, the labor force participation rate for workers aged 65 to 69 was about 33%, and for workers aged 70 to 74 it was less than 20%.⁵³

Another consideration pertaining to the design of the DA factor is that its reliance on older data affects its timeliness. Its basis in a five-year estimate and then its subsequent use for five years means that, in some years, the factor considers data that are more than 10 years old. For example, in the PY2018-PY2022 period data were based on the 2011-2015 reference period. However, the DA factor is not very dynamic. As such, the effects of the lack of timely data tend to be modest. For example, in PY2023 the reference period for the DA factor shifted from 2011-2015 to 2016-2020, but a substantial majority of states had a change of less than 5% in their relative shares of the factor. (See the "Year-to-Year Variation in the Disadvantaged Adults Factor is Limited" section.)

Employment in Areas of Substantial Unemployment Factor and Excess Unemployment Factor

The ASU factor and the EU factor are not clearly aligned with either the more general Adult Activities population or its more specific priority populations. The statute authorizing the Adult Activities program does not emphasize or even mention targeting services to workers in ASUs or workers in areas with otherwise high unemployment rates. The only references in WIOA to the concepts of ASUs and excess unemployment are in the formula provisions.

The service populations specified by WIOA are either general (i.e., the universal access component of the program) or targeted on the basis of certain personal characteristics, such as income level. WIOA's definition of workers with barriers to employment (a population that receives emphasis in states' coordinated planning processes but does not get explicit priority in the provision of services) also emphasizes personal characteristics such as being an ex-offender or long-term unemployed.

Conversely, the ASU and EU factors are based on unemployment rates, which reflect the overall labor markets and may or may not align with the personal characteristics of the WIOA target populations. For example, an area could have high unemployment rates but relatively few low-income workers, as defined by WIOA. Conversely, areas with high concentrations of low-income individuals but low unemployment rates that do not qualify as ASUs would not be reflected by the ASU factor or EU factor.

Analysis of the Formula Factors' Relationship with Other Metrics: Understanding the Factors in More Common Terms

This section analyzes the relationship between the Adult Activities formula factors and more traditional metrics such as poverty and unemployment. The goal of the section is to offer perspectives on the complex and somewhat opaque formula factors by comparing them to more widely understood metrics. This analysis can offer insight as to how the Adult Activities formula factors may favor or disadvantage states with certain characteristics compared to more traditional metrics. This analysis also provides some perspective on how the formula might allot funds differently if the formula factors were based on more standard measures.

States' relative shares of the formula factors generally track with their relative shares of population. For example, California and Texas have large shares of each factor while Vermont and Wyoming have small shares. As such, much of this section emphasizes state-level differences in relative shares between the WIOA factors and traditional metrics (e.g., if a state's relative share of the ASU factor is higher or lower than its relative share of total unemployment).

This section provides four analyses:

• Comparison between the DA factor and persons in poverty. The analysis shows that, relative to a traditional poverty count, states with higher LLSIL thresholds (an income threshold specific to WIOA) tend to have larger relative shares of the DA factors. These states tend to be concentrated in the Northeast and West.
• Comparison between the ASU factor and total unemployment. Compared to their relative shares of total unemployment, states with the highest unemployment rates capture larger shares of the ASU factor. This trend is most prevalent when nationwide unemployment is low.
• Comparison between the EU factor and the ASU factor. As noted in the prior section, the EU factor is typically a subset of the ASU factor. Generally, the EU factor, like the ASU factor, allocates more funding to states with the highest unemployment rates than would a total unemployment metric. The specific relationship between a state's relative share of the EU factor and the ASU factor varies by whether or not an individual state qualifies as a whole-state ASU (unemployment rate above 6.5%) and the overall number of states that qualify as a whole-state ASU.
• Nationwide comparison between the ASU and EU factors and total unemployment. When the national unemployment rate is high, the ASU and EU factors capture a large portion of a higher level of unemployment. When the national unemployment rate is low, the ASU and EU factors capture a smaller portion of a lower level of unemployment.

This section does not formally analyze the relationship between the EU factor and total unemployment at the state level. Some relationships between the EU factor and total unemployment at the state level can be inferred via their corresponding relationships with the ASU factor. The relationship between total unemployment and the EU factor is briefly discussed in the "Summary of Relationships Between States' Total Unemployment, ASU Factor, and EU Factor" section of this report.

Relationship Between the DA Factor and Persons in Poverty

The DA factor considers (1) persons aged 22 to 72 in households with income at or below the poverty threshold and (2) persons from households with incomes above the poverty threshold but below the 70% of the regional LLSIL (see the "Disadvantaged Adults Factor" section for full methodology). This section focuses on the difference between the base measure of persons in poverty and the final DA factor that is used in allocating Adult Activities funds.⁵⁴

Poverty thresholds are the same for the contiguous states and the District of Columbia.⁵⁵ However, the secondary income thresholds used by the DA factor (the LLSIL) vary by metropolitan area and region, meaning that the LLSIL criteria will capture individuals at higher income levels in some parts of the country.⁵⁶ Generally, states with areas subject to higher LLSIL thresholds will capture more individuals beyond the poverty population than states with areas with lower LLSIL thresholds.

In the most recent data, the national total of individuals in the DA factor (about 27.2 million) was approximately 17% higher than the national total of persons in poverty in the applicable age ranges (about 23.3 million).⁵⁷ The distribution of adults from families with incomes below the 70% LLSIL thresholds but above the poverty thresholds is not uniform across states. Generally, in cases where a state's share of adults from families with incomes below the 70% LLSIL threshold exceeds its share of adults from families with incomes below the poverty threshold, the state benefits from the inclusion of the LLSIL measure in the DA factor relative to using only a traditional poverty metric.

The most direct way to measure which states benefit from the LLSIL component of the factor compared to a traditional poverty count is to compare each state's relative share of the DA factor to its relative share of its poverty population.⁵⁸ For example, Maryland's relative share of persons in poverty in the applicable age range is 1.289% and its relative share of the DA factor is 1.414%. Maryland's share of the DA factor is 9.7% greater than its share of persons in poverty, meaning that Maryland's share of the one-third of formula funding that is allotted under the DA factor is 9.7% more than what it would be if the DA factor were replaced by a traditional poverty metric. Each state's relative share of the DA factor that was used to allocate funds in PY2023 compared to its relative share of poverty is depicted in Figure 2.

The states with the largest percentage difference between their relative shares of the poverty count and their DA factor are Hawaii and Alaska, both of which benefit from the DA factor. Hawaii and Alaska have separate, higher poverty thresholds than the other states, but they also have the highest LLSIL levels.⁵⁹ These higher LLSIL levels allow higher numbers of persons to be above the poverty levels but below the 70% LLSIL level, and therefore be captured by the DA factor. In the most recent data, Hawaii's relative share of the DA factor was 31% above its relative share of the poverty count, while Alaska's relative share of the DA factor was 25% above its relative share of the poverty count.

Among the contiguous 48 states and the District of Columbia, states in the Northeast and West tend to have higher relative shares of the DA factor than their relative shares of poverty, while states in the Midwest and South tend to have lower relative shares of the DA factor compared to their relative shares of poverty.⁶⁰ This is because the Northeast and West regions, and metro areas within those regions, tend to have higher LLSIL thresholds, which means those states capture more individuals between the standard poverty income and the area-specific LLSIL.⁶¹

For example, in the 2016-2020 period that will inform allotments for PY2023 through PY2027, Arizona (West region) had a relative share of the DA factor (2.493%) that was about 10% higher than its relative share of the poverty population (2.265%). Conversely, Michigan (Midwest region) had a relative share of the DA factor (2.924%) that was about 8% lower than its relative share of the poverty population (3.193%). Comparisons of the DA factor and poverty estimates for the 2016-2020 period that will inform allocations for PY2023 through PY2027 are in Figure 2.

Figure 2. Percentage Difference Between States' Relative Shares of the DA Factor and Relative Shares of Poverty PY2023 Allocation Data

Source: See Appendix B for data sources and main text for full calculation methodology. Note: Positive percentages generally reflect states that benefit from the DA factor compared to a traditional poverty count.

Relationship Between the ASU Factor and Total Unemployment

The ASU factor counts a state's unemployment in areas with an unemployment rate of at least 6.5% and disregards unemployment in areas with lower unemployment rates. A major issue in assessing the relationship between a state's ASU factor and its total unemployment is whether or not it has a statewide unemployment rate of at least 6.5%. States with unemployment rates of at least 6.5% qualify as whole-state ASUs and 100% of their total unemployment is reflected in the ASU factor. Conversely, the ASU factor in states with unemployment rates below 6.5% will include only a portion (i.e., less than 100%) of the state's total unemployment.

The definition of the ASU factor may suggest potentially diverse outcomes among states with similar statewide unemployment rates below 6.5%: a state with moderate unemployment statewide may qualify for a lower ASU factor than a state with similar unemployment rate through a combination of low unemployment in some areas and high unemployment in other areas. In practice, however, states with similar unemployment rates tend to have a relatively consistent share of their unemployment captured by the ASU factor.⁶² This largely makes the ASU a reflection of each state's relative share of unemployment, with greater weight given to unemployment at higher rates.

Figure 3 depicts each state's statewide unemployment rate (horizontal axis) and its share of total unemployment that was included in the ASU factor (vertical axis). The chart includes more than 450 observations (each state over nine years). In the approximately 20% of cases since PY2015 where a state's unemployment rate was at least 6.5%, the state's ASU factor considered 100% of its total unemployment. This is consistent with statutory policy.

The major insight of Figure 3 is that for the approximately 80% of instances in which a state's unemployment rate was below 6.5%, there was a consistent positive correlation between a state's unemployment rate and its share of total unemployment that was captured by the ASU factor: as the state's unemployment rate increased, the share of its total unemployment that was captured by the ASU factor increased. For example, when a state's unemployment rate was about 4.0%, the share of total unemployment that was captured by the ASU factor centered around 40%. When a state's unemployment rate was about 5.5%, the share of total unemployment that was captured by the ASU factor centered around 90%.

Figure 3. State Unemployment Rate and Share of Unemployment Included in the ASU Factor PY2015-PY2023

Source: CRS analysis. See Appendix B for data sources and main text for full calculation methodology.

As state examples, in PY2023 the ASU factor in California (5.2% unemployment rate) captured about 90% of the state's total unemployment, while the ASU factor in Wisconsin (3.1% unemployment rate) captured about 27% of the state's total unemployment.⁶³

Applied to the formula, the pattern of states with higher unemployment rates capturing a larger share of their total unemployment in their ASU factor means that states with the highest unemployment rates in a given year will typically have relative shares of the ASU factor that are higher than their relative shares of total unemployment and vice versa. To continue the prior examples from PY2023, California had a 14.5% relative share of total unemployment, while its relative share of the ASU factor was 18.8%; conversely, Wisconsin had a 1.4% share of total unemployment but less than a 0.6% share of the ASU factor.

Translated into a ratio, the ratio of California's relative share of the ASU factor to its relative share of total unemployment was about 130% in PY2023.⁶⁴ Conversely, Wisconsin's relative share of the ASU factor relative to its share of total unemployment in PY2023 was about 38%.⁶⁵

Variations of Distribution by Overall Labor Market Conditions

The distribution of states' ratios of their relative shares of the ASU factor to their relative shares of total unemployment varies by the labor market conditions in the reference period. When national unemployment is high and a large number of states qualify as whole-state ASUs, the distribution tends to be narrower and states' relative shares of the ASU factor more closely track their relative shares of total unemployment. In these environments, each state that qualifies as a whole-state ASU in a given year has the same ratio of its relative share of the ASU factor to its relative share of total unemployment (typically just over 100%). Conversely, in cases where overall unemployment is low, a smaller number of states with higher unemployment rates can have much higher ratios (over 150% in numerous cases). In other words, relative to a traditional unemployment metric, the ASU factor allocates funding toward the individual states with the highest unemployment rates more intensively when overall unemployment is low compared to when it is high.

Figure 4 illustrates these principles by presenting the ratio of each state's relative share of the ASU factor to its relative share of total unemployment in PY2015 and PY2020.⁶⁶ In PY2015 (depicted in blue below), the national unemployment rate was 6.8%, and 28 states had unemployment rates of at least 6.5%. In PY2020 (depicted in orange below), the national unemployment rate was 3.8% and one state had an unemployment rate of at least 6.5%.⁶⁷

In PY2015, each of the 28 states that qualified as a whole-state ASU had a ratio of about 104%. In other words, no state's relative share of the ASU factor was more than 4% higher than its relative share of total unemployment. For example, Tennessee was one of 28 states with an unemployment rate of at least 6.5% but its relative share of the ASU factor (2.190%) was only slightly higher than its relative share of total unemployment (2.107%).

In PY2020, 10 states had a ratio of relative share of the ASU factor that was at least 150% of their relative share of total unemployment. In these 10 states, the state's relative share of the ASU factor was at least 50% higher than its relative share of total unemployment. For example, Louisiana's relative share to the ASU factor (2.558%) was 63% higher than its relative share of total unemployment (1.566%).

Figure 4. Relationship between Each State's Relative Share of the ASU Factor and Relative Share of Total Unemployment, by Unemployment Rate PY2015 and PY2020

Source: See Appendix B for data sources and main text for full calculation methodology. Notes: Ratio on the vertical axis is topcoded at 200%. In actuality, the highest ratio was 333% and the second-highest was 218%. Unemployment rate on the horizonal axis is topcoded at 10%. In actuality, the highest rate was 14.3% and the second-highest was 8.9%.

Relationship Between the EU Factor and ASU Factor

As discussed previously, the EU factor is typically a subset of the ASU factor.⁶⁸ A state's EU factor in a given year has distinct relationships with the ASU factor depending on whether or not the state qualifies as a whole-state ASU. The relationship between the EU factor and ASU factor is more consistent in the approximately 80% of cases since PY2015 when a state did not qualify as a whole-state ASU.

Figure 5 presents the share of each state's EU factor as a percentage of its ASU factor by unemployment rate since PY2015 (each dot represents one state's EU factor as a share of its ASU factor for one year). Figure 5 shows two distinct trends:

• In the approximately 80% of cases where a state did not qualify as a whole-state ASU (i.e., it had an unemployment rate of less than 6.5%), the EU factor was typically about 31% of the ASU factor. As discussed previously, the unemployment rate within constructed ASUs is typically close 6.5%. For these states, the EU factor will be calculated as the 2.0 percentage points of the state civilian labor force between 4.5% and 6.5% within the ASU. Mathematically, 2 percentage points is approximately 31% of 6.5%, so the EU factor will be about 31% of the ASU factor.⁶⁹ The close relationship between the ASU factor and EU factor for states that do not qualify as whole-state ASUs underscores the significance of a state constructing its ASUs to maximize its ASU factor and corresponding initial grant.⁷⁰
• In the approximately 20% of cases where a state qualified as a whole-state ASU, the EU factor was at least 31% of the ASU and increased as a share of the ASU as the state unemployment rate increased.⁷¹ For example, in a case where a state's unemployment rate is 6.6%, the state's EU factor will be based on the 2.1 percentage points of the state civilian labor force in excess of 4.5%, or about 32% of its ASU factor. If a state's unemployment rate is 8.5%, the state's EU factor will be based on the 4.0 percentage points of the state civilian labor force in excess of 4.5%, or approximately 47% of the ASU factor.

Figure 5. Relationship Between the ASU Factor and EU Factor, by Unemployment Rate PY2015-PY2023

Source: See Appendix B for data sources and main text for full calculation methodology. Notes: The unemployment rate is topcoded at 10% to simplify presentation. Limited instances in which unemployment rate data were unavailable or a state had an ASU factor of zero are omitted from the figure.

Relationship between the Relative Share of the EU Factor and the Relative Share of the ASU Factor

This relationship between an individual state's relative share of the ASU factor and its relative share of the EU factor varies by overall labor market conditions. In cases where overall unemployment is low and few states qualify as whole state ASUs, each state's relative share of the EU factor correlates closely with its relative share of the ASU factor. In years where overall unemployment is high and many states qualify as whole state ASUs, the states with the highest unemployment rates capture shares of the EU factor that are higher than their relative shares of the ASU factor while states with lower unemployment rates have shares of the EU factor that are lower than their relative shares of the ASU factor.

Figure 6 presents each state's relative share of the EU factor as a percentage of its relative share of the ASU factor in PY2015 (a high unemployment environment) and PY2020 (a low unemployment environment). In cases where the percentage is above 100%, the state's relative share of the EU factor is greater than its share of the ASU factor. In cases where the percentage is less than 100%, the state's relative share of the EU factor is less than its relative share of the ASU factor. If the EU factor were perfectly correlated with the ASU factor, each state's relative share of the EU factor would be 100% of its relative share of the ASU factor.

The orange dots in Figure 6 depict the relationship between states' shares of the EU factor and ASU factor in PY2020 (which had a national unemployment rate of 3.8%). Every state except one had an unemployment rate below 6.5%. For those 51 states, their relative shares of their EU factor as a percentage of their relative shares of the ASU factor clustered around 100%. In other words, for 51 states the relative share of the EU factor was closely correlated to the relative share of the ASU factor.

The blue dots in Figure 6 show the relationship between the ASU factor and EU factor in PY2015 (which had a national unemployment rate of 6.8%, and 28 states qualified as whole state ASUs). In this year, the nine states with the highest unemployment rates had relative shares of the EU factor that were more than 110% of their relative shares of the ASU factor. The states with higher unemployment rates captured a larger share of the EU factor, thereby reducing the relative shares of the EU factor for states with lower unemployment rates. In PY2015, 21 states with unemployment below 6.5% had relative shares of the EU factor that were between 80% and 83% of their ASU factor.⁷²

Figure 6. Relationship between Each State's Relative Share of the EU Factor and Relative Share of the ASU Factor, by Unemployment Rate PY2015 and PY2020

Source: See Appendix B for data sources and main text for full calculation methodology. Notes: Unemployment rate is topcoded at 10%.

The data in Figure 6 demonstrate two potential roles for the EU factor in the allotment formula. In years where unemployment is low (like PY2020), each state's relative share of the EU factor is largely derivative of its relative share of the ASU factor, essentially making the ASU factor double weight in allotting funds. In years where unemployment is higher, the EU factor has a distinct role relative to the ASU factor, allotting more funding to states with the highest unemployment rates and effectively levying a ratable reduction on the EU factor for states with unemployment below 6.5%.

Summary of Relationships Between States' Total Unemployment, ASU Factor, and EU Factor

Table 5 summarizes the direction and nature of the relationships between total unemployment, the ASU factor, and the EU factor that were discussed in the prior sections. The relationships vary based on whether or not a state has an unemployment rate of at least 6.5%.

In two of the four instances, the relationships are fixed: when a state qualifies as a whole-state ASU, 100% of total unemployment is captured by the ASU factor; and when a state does not qualify as a whole-state ASU, the EU factor will typically be about 31% of the state's ASU factor.

In the remaining two instances, the relationship is variable but follows a pattern. When a state does not qualify as a whole-state ASU, the ASU factor is a fraction of total unemployment, which typically increases as the state's unemployment rate approaches 6.5%. When a state qualifies as a whole-state ASU, the EU factor as a percentage of the ASU factor increases as the state's unemployment rate increases.

While the table does not explicitly compare the EU factor to total unemployment, the information in it can be used to determine these trends. When a state does not qualify as a whole-state ASU, the EU factor is typically 31% of a share of total unemployment, which generally increases with the state's unemployment rate. When a state qualifies as a whole-state ASU, the state's EU factor is at least (and usually more than) 31% of the state's total unemployment and the percentage increases as the unemployment rate increases.⁷³

Relationship Between the Aggregate Nationwide ASU and EU Factors and Total Unemployment

The nationwide relationship between the ASU and EU factors and total unemployment also varies by labor market conditions. The national trends are a composite of statewide trends: when overall unemployment is higher, the share of total unemployment that is captured by the ASU and EU factors increases.

In years with higher overall unemployment rates, a larger number of states qualify as a whole-state ASU and therefore have 100% of their state's unemployment considered in the Adult Activities allotment. Conversely, in years with lower overall unemployment, states have to construct more restrictive ASUs, which results in larger shares of total unemployment being excluded from the ASU factor (and therefore the Adult Activities formula) entirely.

Table 6 presents annual data on total unemployment and the share of total unemployment captured by the ASU and EU factors. In the three years between PY2015 and PY2023 when the unemployment rate was at least 5.9%, more than 90% of total nationwide unemployment was considered in states' ASUs factors. In the reference period for PY2020, when unemployment was 3.8%, less than 50% of total unemployment was considered. The EU factor, which is based on a smaller portion of unemployment but has a close relationship with the ASU factor, fluctuated in a similar pattern: higher shares of total unemployment were captured in the EU factor when overall unemployment rates were higher.

In years with lower unemployment rates, the combination of lower total unemployment and a smaller portion of total unemployment considered in the ASU factor means that funds are allocated on the basis of a much smaller population than in higher-unemployment years. For example, in PY2022 the total ASU population was 10.6 million, nearly four times the 2.8 million ASU population in PY2020. An even stronger discrepancy exists with the EU factor where the PY2022 population (about 4.3 million) was nearly five times the PY2020 population (about 873,000). Because a fixed share of funding is allotted on the basis of each formula factor, these variations mean that the share of funding associated with each individual captured by the factor is much greater when unemployment is low compared to when it is high.

Analysis of Year-to-Year Changes in Factor Data: Large Fluctuations in the ASU and EU Factors

This section uses formula data between PY2015 (the effective date of WIOA) and PY2023 (the most recent program year) to analyze the year-to-year changes in formula factors. The analysis focuses on the ASU and EU factors, which are consistently more dynamic than other labor market indicators such as unemployment and civilian labor force. The limited dynamism of the DA factor is discussed in the first section, but it is largely omitted from the subsequent analysis because this factor sees little year-to-year variation.

A central question underlying this analysis is to what degree volatility in formula factors is desirable. Some level of volatility is expected, and is desirable inasmuch as it reflects underlying changes to states' economic conditions and corresponding need to adapt workforce funding. However, the presence of adjustment provisions designed to limit annual change (e.g., hold harmless) suggests that higher levels of volatility are undesirable, even if there are large short-term changes in states' relative economic conditions.

This section describes the relative share of formula factors that are applied to states' initial allotments, prior to the application of the adjustment provisions. The adjustment provisions (described in principle previously and in practice subsequently) somewhat temper some of the large annual swings, though some of the large fluctuations can also have longer-term implications.

Year-to-Year Variation in the Disadvantaged Adults Factor is Limited

Under current practice, the DA factors are only updated once every five years. As such, in four out of every five years, states' relative shares of this factor do not change. Since the effective date of WIOA, these factors have been updated twice: between PY2017 and PY2018 and between PY2022 and PY2023.⁷⁴

Table C-3 presents data on the DA population and relative share for each state and changes when the reference period shifts. When the DA factors are updated, the changes in relative share tend to be modest in spite of the time between updates.

• Between PY2017 and PY2018, 49 of the 52 states had a change in relative share of less than 10%, and within this group 31 states had a change of less than 5%.
• Between PY2022 and PY2023, 45 of the 52 states had a change in relative share of less than 10%, and within this group 36 states had a change of less than 5%.⁷⁵

Year-to-Year Variation in ASU Factor, EU Factor, and Traditional Labor Market Indicators

Figure 7 categorizes states by the percentage change in their relative shares of the ASU factor and the EU Factor. For comparison, the table also includes changes in relative share in two traditional labor market indicators (state-level total unemployment and civilian labor force). The more limited dynamism of the traditional labor market indicators suggests that the volatility of the ASU and EU factors are specific to the design of those factors and not inherent to labor market indicators in general.

The distribution of changes varied by year. The three program years that included data subsequent to the beginning of the COVID-19 pandemic (PY2021 through PY2023) generally had more states with greater year-to-year variations than the prior years, though the volatility of the ASU factor and EU factor exceeded that of the traditional indicators in all years.⁷⁶ Table C-4 includes the underlying data in Figure 7 as well as the average number of states in each category over the eight-year reference period. The subsequent analysis focuses on these averages.

Figure 7. Distribution of Change of Relative Share from Prior Year of Adult Activities Factors and Traditional Labor Market Indicators

Source: CRS analysis. See Appendix B for data sources and main text for full calculation methodology. How to Read Chart: Vertical bars with more blue (especially more light blue) indicate more states with small year-to-year changes. Vertical bars with more turquoise and orange indicate more states with large year-to-year changes. Notes: Limited instances where data were not available or a state's factor was zero are omitted from the analyses, resulting in fewer than 52 states being included in the analysis for a given year. See Appendix B for more details.

Compared to Traditional Labor Market Indicators, the ASU and EU Factors Were Less Likely to Have a Change of Less Than 10%

Over the eight-year period, an average of about 12 states (less than 25%) had a change of less than 10% in their ASU or EU factor. Conversely, an average of 33 states (more than 60%) had a change of less than 10% in total unemployment. Annual data are depicted visually in Figure 7 where the light blue section of the bar (indicating changes of less than 10%) is smaller for the ASU and EU factors than for total unemployment. Civilian labor force was more stable than total unemployment, with only a single instance of a state reporting a year-to-year change of more than 10% during the reference period.⁷⁷

Changes of 30% or Higher Were Much More Likely Among the ASU and EU Factors

Over the course of the reference period, an average of 20 states had a year-to-year change in the ASU factor and an average of 21 states had a change in the EU factor of at least 30%. Conversely, an average of two states had a comparable change in their shares of total unemployment. In other words, changes of at least 30% were about 10 times as common in the ASU and EU factors than in total unemployment.

Prior to the pandemic (up to and including PY2020), the contrast was sharper: an average of 18 states had a change to the ASU factor and an average of 17 states had a change to the EU factor of at least 30%. In the same time period, an average of less than one state had a change in total unemployment of at least 30%.

As noted previously, no state's share of the civilian labor force changed more than 10% from year-to-year in the reference period.

Analysis of the Adjustment Provisions: Frequently Applied to Large Effect

As noted previously, the WIOA Adult Activities formula include provisions to establish minimum grants for small states and other provisions to limit year-to-year fluctuations in states' grants. See the "Adult Activities Adjustment Provisions" text box for a brief summary and the "Adjustment Provisions" section for a more detailed description. The baseline levels for the hold harmless and maximum grant provisions are established in terms of states' relative share of funding, not the grant dollar level.

Since the effective date of WIOA, the adjustment provisions were frequently applied to a large number of states' grants. These adjustment provisions are applied after the calculation of each state's initial grant based on the state's relative share of the three formula factors. The most commonly applied provision was the hold harmless. The minimum grant provision was applied less frequently, and the maximum grant provision was applied somewhat rarely.

The minimum grant provision applied to a small group of states every year and a rotating group of states more intermittently. The minimum grant provision could have a large effect on the grant level for the individual states it applied to but generally reallocated a small amount of overall funding among the states.

The hold harmless provisions were the most frequently applied adjustment provision. The total amount of funding that was reallocated to comply with the hold harmless provision typically exceeded the amount that was reallocated to comply with the minimum grant provisions. Interpreting the common application and effects of the hold harmless is more complex. The hold harmless facilitates stability and prevents individual states from experiencing large year- to-year declines in funding. The provisions also create scenarios in which states' actual grants are somewhat decoupled from their shares of the formula factors that are supposed to determine funding. This decoupling applies to both the states the qualify for the hold harmless and states with grants that are reduced to accommodate the hold harmless.

In some cases where a state had an uncharacteristically high share of the formula factors in one year (not unusual in light of the volatility described in the prior section), the hold harmless could creating a lingering effect. As the state's share of the factors regressed to more typical levels, the hold harmless would be applied in subsequent years, resulting in multiple years of funding that was well above the state's share of the formula factors.

Adjustment Provisions Are Applied Frequently

Table 7 presents the frequency of the application of the adjustment provisions in the Adult Activities formula since WIOA took effect in PY2015. The hold harmless is typically the most commonly applied adjustment provision. The application of the hold harmless somewhat mitigates the influence of the factor volatility discussed in the prior section. This report classifies states where no adjustment provisions applied as receiving calculated grants.⁷⁹

In the nine-year reference period, 12 states received the minimum grant at least once. Four states qualified for the minimum grant every year. Five other states received a minimum grant at least four times, and three other states received the minimum once or twice. In cases where a state received intermittent minimum grants, it was typically during a period of low unemployment. For example, Nebraska received minimum grants for each of PY2015-PY2018, a period during which its unemployment was between 2.9% and 3.7%.

As shown in Table 7, the hold harmless provisions have regularly applied during the reference period, with an average of about 18 states qualifying for the hold harmless each year. With the exception of PY2015, when only 7 states qualified for the hold harmless, at least 12 states have qualified for the hold harmless in each year. The hold harmless (which limits decline from the prior year) is applied more frequently than the maximum grant (which limits increase from the prior year), likely because the change threshold for triggering the hold harmless (-10%) is less than the threshold for triggering the maximum grant (+30%).

The frequent application of the hold harmless provisions is likely related to the volatility of some of the formula factors. For example, the prior discussion noted that it is common for two of the three factors to vary in excess of 30% from one year to the next. In cases where the relative share of factors declines, the hold harmless can mute the effect of downward year-to-year changes. In cases where the relative share of factors increases, the maximum grant can limit changes, though the threshold to trigger this provision is higher than the threshold to trigger the hold harmless provision.

The maximum grant provisions were less commonly applied than the hold harmless provisions. In PY2021 (the first year in which the COVID-19 pandemic labor market was included in the reference data), nine states were limited by the maximum grant provision. No more than three states received a maximum grant in any other year and there were three years in which no state was limited by the maximum grant provisions.

Quantifying the State-Level Effects of the Adjustment Provisions

Compared to what states would have received purely on the basis of their relative share of the formula factors, the adjustment provisions increase some states' grants and decrease the grants for others. Specifically, relative to calculated grants, the adjustment provisions increase grants for states that qualify for a minimum or hold harmless and reduce grants for states that qualify for a maximum grant or are not subject to any of the adjustment provisions.⁸⁰ To quantify the effect of the limiting factors, CRS compared each state's share of funding to its average share of the three formula factors, which would have determined each state's grant in the absence of the adjustment provisions. For example, in PY2023, Virginia's relative share of the formula factors was 1.202%. However, Virginia's relative share of funding in the prior year was 1.716%, which meant that the lowest share of funding it could receive, per the hold harmless provision, was 1.545% of total funding (90% of 1.716%). The state's relative share of funding (1.545%) was 28.5% higher than its relative share of the formula factors (1.202%), so Virginia's effect from the adjustment provisions was +28.5%.

Table 8 depicts each state's final grant relative to its share of the formula factors in PY2023. Positive effects are generally associated with the minimum grant provisions and hold harmless provisions. In other words, because of the adjustment provisions, these states' relative share of funding is typically higher than their relative share of the three formula factors.⁸¹ Negative effects are associated with the maximum grant provisions and states that were not subject to an adjustment provision.

All states with calculated grants have the same percentage difference between their grants and their shares of Adult Activities funding in PY2023 (-6.3%). This uniform decrease reflects the ratable reduction that is applied to states to accommodate the hold harmless and minimum grant provisions.⁸²

Table C-5 presents similar data for each year since WIOA took effect in PY2015. In each year, the ratable reductions to comply with the minimum grant and hold harmless provisions were greater than the ratable increases to comply with the maximum grant provisions. As such, the relative shares of funding for states with calculated grants were lower than those states' relative shares of the formula factors.

In each year, the share of grant funding for each state with a calculated grant was a uniform percentage below the state's relative share of the formula factors. In PY2023, this percentage was 6.3%. The highest uniform reduction since WIOA took effect was in PY2022, when states with calculated grants had shares of funding that were about 8.4% less than their shares of the formula factors. The smallest uniform reduction was 1.6% in PY2015.

Effect of Minimum Grant Provisions

The minimum grant provisions typically provide for some of the largest percentage increases under the adjustment provisions. For example, in PY2023 Vermont's relative share of the three factors was 0.07%. The minimum grant provisions increased the state's share of total funding to 0.25%, an increase of almost 270%. While states that were subject to the minimum grant tended to have larger increases relative to their share of the formula factors, states that received the minimum grant accounted for a relatively small share of total funding. In FY2023, the seven states that qualified for the minimum share of 0.25% of total funding accounted for 1.75% of total grants. These states collectively had about 0.9% of the total formula factors, meaning that the minimum grant provision, on average, almost doubled their grants but also resulted in the reallotment of less than 1.0% of total funding.

Effect of Hold Harmless Provisions

The effect of the hold harmless on individual states' grants varies. In PY2023, 23 states were subject to the hold harmless. Of these, 12 received a relative share of funding that was at least 20% greater than their relative share of the formula factors, and 4 of the 12 received a relative share of funding that was at least 40% greater.

Five states that were subject to the hold harmless received grants that were less than their relative share of the formula factors, but the difference was less than the -6.3% for the states that received a calculated grant. Without the hold harmless provision, these five states would have been subject to a full ratable reduction that would have pushed their relative share of funding below 90% of their relative share from the prior year. The hold harmless effectively stopped the ratable reduction in these states and prevented them from having their grant amounts reduced by the same amount as the reduction for calculated grant states.

Hold Harmless in Consecutive Years

Table 9 divides hold harmless states by the number of consecutive years the state has qualified for the hold harmless. In many cases, states with the largest effect from the hold harmless are those that are receiving the hold harmless in consecutive years. For example, in PY2023, of the 12 states where the hold harmless increased their relative shares of funding by at least 20%, 10 were receiving the hold harmless for at least a second consecutive year. Similar trends exist in prior years. For example, in PY2018 there were nine states where the hold harmless increased their relative shares of funding by at least 20%, and eight of these states were receiving the hold harmless for at least the second consecutive year.

This phenomenon of the hold harmless being applied in multiple years may be partially related to the dynamic nature of some of the formula factors. As described in the prior section, it is not unusual for a factor to change by 50% or more from one year to the next. In some of these cases, a state receiving a hold harmless in consecutive years can be a residual effect of a single prior year in which the state had an unusually high share of the formula factors. For example, in PY2016 Indiana's relative share of the formula factors was 1.945% and the state received a calculated grant. In PY2017, the state's share of the factors declined to 1.650% (-15.2%), but the hold harmless limited its change in funding share to -10%, or 1.751% of grant funding. In the subsequent PY2018, the state's share of the formula factors declined further to 1.270%, a decrease of 23% from the prior year, but the hold harmless amount was based on the prior year's hold harmless allotment, meaning that the state received 90% of its relative share from the prior year, or 1.575% of total funding. Between PY2016 and PY2018, Indiana's share of factors declined about 35%, but due to the repeated application of the hold harmless provisions, the state's share of funding only declined 19%. The controlled decline of the state's funding in PY2017 and PY2018 also impacted the state receiving a hold harmless again in PY2019 when its relative share of formula factors declined further.

Effect of Maximum Grant Provisions

The maximum grant provisions reduce the grants for the applicable states and increase grants for states with calculated grants. Typically, relatively few states are affected by the maximum grant provisions. In the nine years since WIOA took effect, there were zero states or one state with a maximum grant in six years. PY2021 was a major exception, with nine states qualifying for a maximum grant. PY2021 was the first year in which data from the COVID-19 pandemic were considered in the allotment formula.

The application of the maximum grant provision does create a ratable increase in funding for other states, but in each year this increase was more than offset by the ratable reductions to accommodate the minimum grant and hold harmless provisions.

Calculated Grants

Calculated grants reflect the combination of ratable reductions to accommodate the minimum grant and hold harmless provisions and ratable increases following the application of the maximum grant provisions. States with calculated grants consistently have a lower relative share of grant funding compared to their relative share of the formula factors. In PY2023, the relative share of grant funding for each state that received a calculated grant was about 6.3% less than its relative share of the formula factors.

Aggregate Effects of Adjustment Provisions

The aggregate effect of each limiting provision can be estimated by comparing the total share of the formula factors of a group of states subject to a given provision to the total share of the grant funding for states in each limiting provision category. These calculations for PY2023 are presented in Table 10. Generally, the minimum grant provisions and the hold harmless provisions can be seen as raising the grant for the states benefitting from those provisions at the cost of reducing grants for other states.

The 23 states subject to the hold harmless in PY2023 had 27.6% of the factors but received 31.4% of the funding. This amounts to an average increase of 13.6% for the 23 states relative to what they would have received in the absence of such a policy. Reallocations due to the hold harmless were the primary driver of the uniform 6.3% difference between the relative share of formula factors and the relative share of grant funding for states that were not subject to any of the adjustment provisions (i.e., states receiving calculated grants).

Reallotments for small state minimums tended to have a larger effect on the states receiving the minimums but smaller effects on the states losing funds. In PY2023, states subject to the minimum grant received 1.75% of total funding relative their 0.92% relative share of the factors. This accounted for large increases for some of the states that received the minimum but reallocated less than 1.0% of total funds.

Table C-6 presents the relative share of formula factors and grants (similar to Table 10) for each year since the effective date of WIOA. While magnitudes vary, several trends are present in most years:

• States subject to the minimum grant provision tend to have the largest percentage differences between their shares of the factors and their shares of funding. However, the reallocation of funding to accommodate the minimum grants is typically 1.0% or less.
• The adjustments for the hold harmless provision are typically the largest adjustments in terms of the share of formula funding that is reallotted, though the percentage varied by year. The share of funding that is reallotted under these provisions is typically related to the number of states that qualify for the hold harmless. In the nine years since WIOA took effect, the hold harmless reallocations have accounted for more than 5.0% of formula funding twice and less than 1.0% twice.
• The adjustments for the maximum grant are typically small and facilitate a small ratable increase to states not subject to the adjustment provisions. The maximum grant provisions are applied less frequently than the hold harmless provisions, due at least partially to the higher change threshold (30% increase to trigger maximum grant versus a 10% decline to trigger hold harmless).
• The difference between the relative share of formula factors and grant funds for states that are not subject to adjustment provisions (i.e., calculated grant states) tends to be inversely related to the number of states subject to the minimum grant provision and (especially) the hold harmless. In each year, each state that had a calculated grant was subject to a uniform reduction. For example, in PY2015 seven states qualified for a hold harmless and states with calculated grants had relative shares of funding 1.6% below their relative shares of formula factors. Conversely, in PY2020 20 states qualified for a hold harmless and states with calculated grants had relative shares of funding that were 6.6% less than their relative shares of formula factors.

Summary of Key Findings

The analysis in this report has identified a number of issues with the Adult Activities formula that Congress may choose to consider when reauthorizing WIOA. Many of these issues arise from the statutorily specified formula factors, and include the following:

• The statutory definitions of the formula factors necessitate complex calculation procedures. These procedures are not easily understood and the resulting factors may not be optimal for allocating workforce funding. The DA factor is based on a series of calculations, including one that utilizes an income indicator that has not been fully updated since the 1980s. In the 80% of cases since PY2015 where states constructed their ASUs, both the ASU and EU factors were largely based on the ability of state agencies to strategically identify contiguous areas within the state with a collective unemployment rate of at least 6.5%.
• The ASU and EU factors do not clearly align with Adult Activities populations. Adult Activities is a universal access program that gives priority to low-income workers and other disadvantaged populations. The ASU and EU factors, which determine two-thirds of initial state grant allotments, are calculated on the basis of concentrations of unemployment, which vary from year-to-year and may not reflect the populations that the WIOA statute emphasizes.
• Relative to more traditional indicators, the formula factors favor states with certain characteristics. Compared to a traditional poverty indicator or another indicator that uses a nationwide income threshold, the DA factor uses a regional metric that favors states with higher costs of living. The ASU and EU factors capture larger shares of total unemployment in states with higher unemployment rates and therefore, relative to a traditional unemployment metrics, allocate more funding to states with higher unemployment rates.
• The ASU and EU factors are typically more volatile than traditional metrics. The majority of states' relative shares of the ASU and EU factors frequently vary by more than 30% from one year to the next. Traditional labor market indicators such as total unemployment and civilian labor force are typically less volatile.
• The adjustment provisions have significant effects on final grant levels. The adjustment provisions (particularly the hold harmless) temper some of the factor volatility but also result in grant levels that are decoupled from states' shares of the formula factors. The ratable reductions to accommodate the adjustment provisions typically lead to grants for some states that are well below their shares of the formula factors. This is a common issue in formula grant programs where the desire to have enough stability in year-to-year funding to operate stable programs is balanced against allocating funds in a manner that aims to capture shifts in the formula factors.

Potential Policy Options

Congress may consider modifying the Adult Activities formula as part of WIOA reauthorization. This section discusses advantages and drawbacks of two formula factor frameworks. It also discusses policy options to limit large changes in grant amounts that could result from the implementation of new factors or other changes.

Potential New Factor Strategies and Associated Considerations

Potential new factors would likely fall into two categories: (1) traditional indicators and (2) targeted indicators. Generally, traditional indicators will be more transparent and less volatile than the current indicators, while targeted indicators can be more precisely aligned with WIOA priority populations. In all cases, adjustments to the weight of factors could increase or decrease the influence of certain metrics.

Traditional Indicators

Traditional indicators would generally include data that are regularly published by major federal statistical agencies (and are used for comparisons throughout this report). These could include indicators such as measures of the civilian labor force, total unemployment, and persons in poverty.

Compared to the current Adult Activities formula factors, the primary advantages of these types of formula factors are transparency and stability. These indicators are commonly understood and developed using widely accepted methodologies. As demonstrated in the prior sections of this report, traditional indicators are also subject to less volatility than the current ASU and EU factors.

A potential criticism of these indicators is that they are not fully aligned with the WIOA population. Indicators like civilian labor force and total unemployment offer a sense of the scale of a potential WIOA population but do not, for example, include individuals who are out of the labor force and could benefit from WIOA services. Further, similar to the current ASU and EU factors, most traditional indicators reflect the economic conditions of a place rather than the personal characteristics that the Adult Activities program targets.

Targeted Indicators

Another option could be to allot Adult Activities funding using factors that are aligned with the WIOA priority populations. These could include selected existing indicators that could serve as a proxy for target populations (such as recipients of certain public benefits) or constructed indicators that try to more comprehensively count members of priority population criteria.

Similar to the existing DA factor, a constructed indicator could be developed to count individuals that meet a specified set of criteria. For example, existing survey data could be used to develop estimates of a specified population, such as the WIOA priority populations, perhaps using a multistage process similar to that used by the DA factor.⁸³

The major advantage of using targeted factors is that they could be more closely aligned with each state's relative share of a specific population. Rather than indicating the conditions of a place, a targeted indicator could focus on the personal characteristics that the Adult Activities program explicitly targets.

One potential drawback of a new targeted factor is that the construction would be complex and not transparent. This could effectively be seen as trading one opaque factor for another. If the factor includes a number of subpopulations, it may not be possible to easily determine what elements are driving year-to-year changes.

Another potential drawback of a new targeted factor is that the complexity could lead to issues that are similar to those currently seen in the existing DA factor, including being constructed from multiple years of data and not being updated each year.

The volatility of a potential new targeted factor could depend on its construction. If it is largely population-based, it may be relatively stable, similar to the existing DA factor. However, if it relies on more dynamic elements, such as employment status, it may fluctuate from year-to-year similar to the existing ASU and EU factors.

Weighting Considerations

Decisions to modify or retain the existing formula factors may be accompanied by considerations of the appropriate weight for each factor. Currently, the three formula grants in Title I of WIOA each have three equally weighted formulas.

Weighting could be changed by increasing or decreasing the number of factors but retaining equal weights for each factor. It could also be possible to design a formula in which some factors have greater weight than others. For example, a three-factor formula could allocate 50% of funding on the basis of one factor that legislators wish to emphasize and allocate 25% of funding on the basis of each of two other factors.

Potential Policy Options to Facilitate Continuity with the Existing Formula and Allotments

Changing the formula factors would likely be advantageous to certain states and reduce funding for others. These dynamics could result in changes to the formula being considered controversial and could require states to make changes to their workforce system operations and strategies. Should Congress opt to make changes to the formula, it could consider a variety of approaches to limit the effects of formula factor changes, including the following:

• Construction of adjustment provisions. Hold harmless and maximum grant provisions can limit the immediate changes that result from formula changes. For example, the current hold harmless provisions would limit any decline in a state's share of funding to 10% from the prior year. A potential related approach may be to create a separate hold harmless provision based on 100% of a prior year's funding level to ensure that no state's funding would decline in absolute dollar terms after formula refinements were adopted. This could become a new foundation that is used every year or it could decline over time.
• Choosing new factors that have some alignment with prior factors. Partial continuity with the existing formula factors can inform the selection of new factors. For example, persons in poverty is a key element of the DA factor and if Congress were looking for a more transparent factor that had significant alignment with the existing factor, it might consider replacing the DA factor with a count of persons in poverty. This approach could be seen as trying to maintain some of the principles of the current formula while moving to factors that are more transparent.
• Phasing in a new formula. To limit the immediate effects of a formula change, a portion of a program's funds could be allocated using the new formula and the remainder allocated using the old formula during a phase in period. For example, it may be possible to structure a change so that in the first year 80% of the funding would be allotted using the prior formula and 20% allotted using the new formula. The percentage allotted using the new formula could increase 20% each year, meaning that by the fifth year, 100% of funds would be allotted using the new formula.
• Increase total funding for grants. Any of the aforementioned strategies could be combined with an increase in total funding, if such an increase was deemed desirable by Congress. Such an approach could increase funding for most or all states and reduce possible declines in funding among states with shares of the new factors that are smaller than their shares of the current factors. An increase in overall funding could mean that even if a state's relative share of new factors were less than its relative share of previous factors, the state's overall funding may still increase.

Appendix A. Historical Background and Legislative History

This section provides a legislative history of federal workforce development statutes, with an emphasis on placing the current Adult Activities formula in a historical context. The current Adult Activities formula in WIOA generally follows the approach of a similar program in the Workforce Investment Act of 1998 (WIA; P.L. 105-220), which largely follows the approach of a similar program in the Job Training Partnership Act of 1982 (JTPA; P.L. 97-300). CRS's review of the legislative history of WIOA, WIA, and JTPA did not reveal a great deal about the congressional intent behind the development of the formula and subsequent decisions to retain it.

Table A-1 depicts the major formula elements of grants in each of WIOA, WIA, and JTPA. This review focuses on the formula elements of the relevant programs established in statute and only notes major changes in other programmatic details. The original JTPA formula grant initially targeted both disadvantaged adults and disadvantaged youth. It was split into two programs by the Job Training Reform Amendments of 1992 (P.L. 102-367), which created separate programs for youth and adults. Only the latter is included in Table A-1.

Job Training Partnership Act of 1982

JTPA marked a significant change from prior federal workforce legislation for disadvantaged workers. Prior to JTPA, the primary federal workforce development legislation was the Comprehensive Employment and Training Act (CETA), which was established in 1973 and amended several times prior to the enactment of JTPA. The largest component of CETA was subsidized public service employment, typically facilitated by local "prime sponsors," which were usually units of local government. JTPA marked a concerted departure from CETA, increasing the role of the private sector, emphasizing training over subsidized employment, and expanding roles for states and governors rather than localities.⁸⁴

In the 97^th Congress, separate bills to replace CETA were reported by the relevant committees in the House and Senate. Each bill included a formula grant to states to provide job training for economically disadvantaged individuals:

• H.R. 5320, as reported, had a formula with four equally weighted factors: (1) the number of unemployed persons, (2) the excess number of unemployed, (3) the number of unemployed persons living in areas of substantial unemployment, and (4) the number of adults in low-income families in the jurisdiction. These factors were similar to the Transitional Public Service Employment program formula in CETA (one of several formula grants in that law), but the accompanying committee report did not note the similarity or otherwise establish a rationale for the factors.⁸⁵
• S. 2036, as reported, had a formula with two equally weighted factors: (1) the number of economically disadvantaged persons in the labor force and (2) the number of long-term unemployed. The bill also established a minimum grant of 0.25% and a hold harmless equal to 90% of the state's allotment percentage from the prior year. The accompanying committee report did not explain a rationale for these factors or adjustment provisions.⁸⁶

The subsequent bill approved by a conference committee established a formula grant program under Title II-A of JTPA. The formula generally retained three of the four factors from H.R. 5320 and specified that funds would be allocated on the basis of these three equally weighted factors. The Senate provisions related to minimum grants and a hold harmless were also retained. The conference report did not offer a rationale for these decisions.⁸⁷

The definitions of the unemployment in areas of substantial unemployment factor and excess unemployment factor in the final JTPA statute largely mirror the current definitions for the ASU and EU factors in WIOA.⁸⁸ The disadvantaged individuals factor in JTPA is different from the current DA factor in that it does not have an age restriction. It is similar to the current DA factor in that it considers persons with incomes below the poverty line or 70% of the lower living standard income level.

JTPA Amendments of 1992

The JTPA Amendments of 1992 (P.L. 102-367) split the Title II-A program in JTPA into two separate programs that separately targeted disadvantaged adults and disadvantaged youth. The two formula programs had two similar factors (ASU and EU) as well as third factors that were related but distinct (disadvantaged youth and disadvantaged adults). For the adult program, the law established the current 22-72 age range that is considered by the current DA factor as well as the policy of excluding college students and members of the Armed Forces.⁸⁹ The law also established the current maximum grant of 130% of a state's relative share from the prior year.

Workforce Investment Act of 1998

The 105^th Congress enacted WIA, which authorized appropriations for each of FY1999 through FY2003. WIA established an Adult Employment and Training Activities formula grant program for states. Various elements of the program differed from the Disadvantaged Adult program in JTPA, but the formula had the same factors and adjustment provisions as the Adult Training Program under the JTPA Amendments of 1992.⁹⁰

During the development of WIA, bills were reported by the relevant committees in the House and Senate. Both bills retained the same formula from JTPA for the revised adult program. Both committee reports noted this similarity.⁹¹ The House report noted that a partial rationale for retaining the JTPA formula was "protecting against major funding shifts among States."⁹²

Selected Legislative Efforts between WIA and WIOA

The authorization of appropriations under WIA expired after PY2003. Prior to the enactment of WIOA in 2014, there were several efforts to reauthorize WIA, including bills that passed either the House or Senate. Some of the bills would have modified the formula for the adult-focused program. This section discusses some unenacted proposals that would have altered the Adult Activities formula. It is not a comprehensive review of efforts to reauthorize WIA.

In some cases, proposed formula changes were part of a broader effort to consolidate formula programs. For example, in the 108^th Congress, the House passed a bill that would have combined the adult program with other programs into a single formula grant. The consolidated program would have distributed a majority of funding under a new formula that would have allotted 60% of the funding on the basis of each state's relative share of total unemployment. Smaller portions of the funding would have been allocated on the basis of states' shares of disadvantaged adults, excess unemployment, and civilian labor force. The bill would have eliminated the ASU factor. Several mechanisms were in place to limit immediate changes under the new formula.⁹³

Other bills retained a dedicated stream for adult activities, but modified the allocation factors. For example, in the 109^th Congress the Senate passed a WIA reauthorization bill that would have retained a separate adult-targeted program but would have modified the formula so that 40% of funds would have been allotted under the ASU factor, 35% would have been allotted on the basis of the DA factor, and 25% would have been allotted on the basis of shares of the civilian labor force (a new factor).⁹⁴

Workforce Innovation and Opportunity Act of 2014

WIOA was enacted in July 2014 and retained many of the principles and administrative structures of the predecessor WIA statute. The Adult Activities formula grant in WIOA had the same formula factors and procedures as the similarly named grant in WIA.

During the development of WIOA, some formula-related changes were considered in the context of broader consolidation, but they were not enacted. In March 2013, the House Education and Workforce Committee reported H.R. 803, which would have consolidated a number of workforce programs (including the Adult Activities program) into a single formula grant to states. The formula would have allocated funds on the basis of four equally weighted factors: (1) unemployment in ASUs (existing Adult Activities formula factor), (2) civilian labor force, (3) long-term unemployment (at least 15 weeks), and (4) disadvantaged youth. The bill would have established a three-year transition period to moderate the effects of the new formula.⁹⁵

In November 2013, the Senate Committee on Health, Education, Labor, and Pensions reported S. 1356 without a written report. The bill would not have consolidated any of the major formula grant programs nor did it propose any formula changes to the adult program.

The final version of WIOA followed the Senate bill and did not consolidate any of the major formula grant programs or change formulas.⁹⁶ The adult program was renamed Adult Activities and retained the formula from WIA.

Appendix B. Data Sources

This report uses a combination of publicly available data and data that were obtained directly from the Bureau of Labor Statistics (BLS).

DOL publishes state-level formula factor data on its website. Data on disadvantaged adults, unemployment in areas of substantial unemployment, and excess unemployment were accessed directly from this site. DOL also publishes details and data on the components of the DA factor. Specific links are provided in the sources for Table B-1.

Data on state-level civilian labor force and total unemployment that were comparable to the ASU and EU factor data were obtained directly from BLS. BLS publishes state-level civilian labor force and unemployment data on a monthly basis. Data are typically published on the third Friday of the month following the reference month.⁹⁷ For example, June 2021 data were published on Friday, July 16, 2021.⁹⁸ The initial published data are preliminary and are revised for up to five years. The data on the BLS website reflect the most recent revisions and not the earlier versions that were used in the creation of the formula factors.

The ASU and EU factor data are calculated on the basis of the state-level unemployment data as of the preliminary June data in the reference year. Subsequent revisions are not considered in the ASU and EU calculations. To ensure comparability with the ASU and EU factors, the state-level civilian labor force and unemployment data as of the preliminary June data for each year are used in this report. For example, the unemployment data for PY2022 are the data for July 2021 through June 2022 as of the initial publication of the June 2022 data.

BLS provided CRS with each state's data for each relevant program year as of the preliminary June data. It was necessary to obtain this data directly from BLS because, as noted previously, data on the public BLS website have been revised and would not reflect the data that were used for the ASU and EU calculations. BLS also provided CRS with unpublished data on the civilian labor force in each state's ASUs and the number of ASUs in each state.⁹⁹

Table B-1 provides sources for the underlying data that were used for the analysis in this report. Other indicators that were calculated by CRS on the basis of these data are not included in the table. These other indicators include traditional metrics such as unemployment rate (calculated on the basis of BLS-provided civilian labor force and unemployment data) and more formula-specific indicators, such as the percentage of total unemployment considered in the ASU factor or states' relative shares of formula factors.

Zero and Missing Data

For the data used in this report, there are limited instances of zero and missing data.

In the reference periods for PY2019 and PY2020, the state of Vermont had an ASU factor and an EU factor of zero. This issue led to Vermont being excluded from some analyses in this report.¹⁰⁰ For example, because it is not possible to calculate a change from a base period of zero, no change was calculated for Vermont's ASU and EU factors for PY2019 to PY2020 or PY2020 to PY2021. This exclusion is why, for example, the bars for the ASU factor and EU factor in Figure 7 do not include all 52 states for the periods ending in PY2020 and PY2021. The period ending in PY2019 was classified as a decline of 100%.

In the reference period for PY2021, BLS was unable to estimate the civilian labor force for Puerto Rico. While this did not affect the formula directly, it did affect some of the traditional labor market indicators that are used for comparison throughout the report. The lack of civilian labor force data for PY2021 means it was not possible to calculate the unemployment rate for Puerto Rico in PY2021, so CRS did not include PY2021 data for Puerto Rico in analyses that consider unemployment rate. The analyses of changes in civilian labor force excluded Puerto Rico for the period ending in PY2021 (because reporting a decline of 100% would be misleading) and for the period ending in PY2022 (because there were no data for the PY2021 base period).

BLS certified Puerto Rico as a whole-state ASU in that reference period. For PY2021, CRS used the published ASU factor for Puerto Rico for the state's total unemployment.

Appendix C. Additional Data Tables