Overview

This document provides an overview of the analyses conducted in developing and evaluating the statistical adjustment models that will be used for Program Years (PYs) 2026 and 2027. The statistical models are used to inform performance negotiations and to assess state performance after each program year, in accordance with WIOA performance accountability requirements. These models apply to the following indicators for WIOA Title I and Title III programs: Employment Rate in the Second Quarter after Exit (ERQ2), Employment Rate in the Fourth Quarter after Exit (ERQ4), Median Earnings in the Second Quarter after Exit (MEQ2), Credential Attainment (CRED), and Measurable Skill Gains (MSG) for the Adult, Dislocated Worker, Youth, and Wagner-Peyser programs. (Note: the CRED and MSG indicators do not apply to the Wagner-Peyser program.)

This report includes the following sections:

  • Modifications - details on the changes made to the this version of the models
  • Prediction Plots - plots showing how the PY 2026-2027 predictions compare to actual PY 2024 performance
  • Final Estimates - the final model estimates (i.e., the coefficients) for PY 2026-2027
  • Final Predictions - the final model predictions used in PY 2026-2027 (i.e., Pre-Program Year Performance Estimate)
  • Full Variable Table - a table showing which variables are in included in each model

Background

The initial methodology of the statistical adjustment models was developed by the U.S. Department of Labor’s Chief Evaluation Office in 2016. The models recommended at that time were implemented as a preliminary test beginning with PY 2017 using WIA data and proxy variables where necessary. Subsequently, the models have undergone revisions and enhancements with each negotiations cycle. For more information on previous versions of the models, see the corresponding Model Summary Reports for those cycles on the State Performance Negotiations Resource Archive or click on the reports that are linked below.

  • WIOA Statistical Adjustment Model Methodology Report - the initial proposed methodology for the WIOA Statistical Adjustment Models. This report provides background on the WIOA requirement of a statistical adjustment model, program and data limitations, a framework for identifying an appropriate model, and a recommended approach to be implemented for future WIOA program years.
  • PY 2020-2021 Model Selection Report - the description of the models chosen for Program Years 2020-2021. In addition to the final model estimates and specifications, this report details the main modifications made at that time which included: how model performance is affected by the data (WIA vs. WIOA), refinement of the model specifications, development of the Measurable Skill Gains models, and other changes.
  • PY 2022-2023 Model Selection Report - the description of the models chosen for Program Years 2022-2023. In addition to the final model estimates and specifications, this report details the main modifications made at that time which included: changing how the economic conditions data is aligned to the WIOA participant data to better measure the effects of the economy on performance, implementing the use of random sample groups of participant data for the Measurable Skill Gains models, adding the Employment Rate 4th Quarter after Exit and Credential Attainment models, and other changes.
  • PY 2024-2025 Model Selection Report - the description of the models chosen for Program Years 2024-2025. In addition to the final model estimates and specifications, this report details the main modifications made at that time which included: the addition of an average wages of labor force variable to the Median Earnings 2nd Quarter after Exit models to account for wage increases over time, a process for normalizing all non-percentage variables in all models, and changes in how the race category variables are defined.

The models for Program Years 2026-2027 have been further refined and this document explains the changes that have been made for this negotiations cycle and provides the final model specifications and estimates. The updated models use the reported WIOA data from PY 2019-2024. For in-depth details of the changes made to these versions of the models see the Modifications tab.




Modifications

For PY 2026-2027, two modifications were made to the industry share variables used in the WIOA Statistical Adjustment Models (SAM). The first modification changes which industry categories are included in the industry share calculation. The second modification introduces an imputation process to address suppressed employment values in the QCEW data. Each modification is described in detail below.


Changed the Industry Share Variables Included in the Models

The industry share variables used in the model are calculated using data from the Bureau of Labor Statistics’ (BLS) Quarterly Census of Employment and Wages (QCEW). These data include monthly employment counts for the following high-level industry titles: Natural Resources and Mining; Construction; Manufacturing; Information Services; Financial Services; Professional and Business Services; Educational and Health Care; Leisure and Hospitality; Other Services; Public Administration; Trade, Transportation, and Utilities; and Unclassified. Employment counts for these industry titles are broken out by the following ownership levels: Private, Federal, State, and Local.

For the SAM, the employment counts across ownership levels are summed by industry. The total count for each industry is then divided by the total employment count across all industries to calculate each industry’s share of total employment.

In previous versions of the SAM, all of these categories were used. However, starting with the PY 2026-2027 models, the Unclassified category is no longer included. This decision was made because that category is infrequently populated and is only used temporarily until employment can be classified into the appropriate industry title. Including it in the industry share calculation introduces noise without meaningful economic signal. As a result, industry shares for the remaining categories are calculated without the Unclassified counts.

In addition, alongside the removal of the Unclassified industry share, the base variable that is dropped from the model was also changed. Dropping one industry share variable is necessary to prevent perfect multicollinearity, since the industry shares across all categories sum to one. In previous versions of the SAM, the Unclassified category effectively served this role by being excluded from the model. In the PY 2026-2027 models, the Other Services variable is explicitly used as the dropped base variable. All other industry share variables are included as predictors in the WIOA SAM, and their coefficients are interpreted relative to the Other Services share.


Added Imputation for Suppressed Industry Share Data

The second modification to the industry share variables addresses data quality. Analysis of the two most recent available quarters of QCEW data (CY 2025 Q1 and Q2) revealed an increase in suppressed values due to data collection challenges stemming from the Federal government shutdown in 2025. BLS suppresses employment counts at the ownership-industry level for several reasons, including privacy concerns (when counts are low), data collection disruptions (e.g., a natural disaster affecting the locality during the reference period), or other data quality issues. This uptick in suppressed values had the potential to impact SAM estimates and adjustments.

In recent years, approximately 10% of all QCEW observations used by the WIOA SAM have been suppressed; however, the vast majority of these suppressed values would have been zero or near zero.

The tables below show the extent to which suppressed values appear in the QCEW data. Suppression rates are higher in certain industries, such as Natural Resources and Mining, Manufacturing, and Construction. However, the large majority of suppressed values fall under State or Local ownership. This limits the overall impact since, for most industries, Private ownership accounts for the majority of total employment in the industry share as used by the SAM.

Prevalence of Suppressed QCEW Data
Overall
Overall Prevalence of Suppressed Values
Total Records Suppressed Count Suppressed Percentage
231300 25422 10.99%
By Industry
Prevalence of Suppressed Values by Industry
Industry Total Records Suppressed Count Suppressed Percentage
Construction 16062 2790 17.37%
Educational and Health Care 26190 264 1.01%
Financial Services 23958 3351 13.99%
Leisure and Hospitality 24792 1842 7.43%
Manufacturingg 13332 2490 18.68%
Natural Resources and Mining 11820 3366 28.48%
Other Services 22782 3723 16.34%
Professional and Business Services 25368 2190 8.63%
Public Administration 19728 324 1.64%
Trade, Transportation, and Utilities 23940 1884 7.87%
iInformation Services 23328 3198 13.71%
By Ownership
Prevalence of Suppressed Values by Ownership
Ownership Total Records Suppressed Count Suppressed Percentage
Federal 54594 0 0.00%
Local 61488 9225 15.00%
Private 65856 582 0.88%
State 49362 15615 31.63%


In past versions of the WIOA SAM, suppressed values were simply treated as zero. This approach worked reasonably well for most suppressed values, since they typically occur in small ownership-industry observations that do not meaningfully affect the total industry count. However, it was discovered that some suppressed values are large enough to distort the industry share calculation when replaced with zero. As discussed above, this issue is particularly pronounced for the two most recent quarters of QCEW data, which are preliminary and therefore have a higher rate of suppression on average. To account for cases where larger values may have been suppressed, and to establish a process for handling potential data disruptions in the future, an imputation process for these data was developed.

Following the discovery of the potential impact of suppressed QCEW data, several options for minimizing the effect of suppression on the SAM were researched and tested. As a result, an imputation process for suppressed values was developed and is applied in the PY 2026-2027 models. The remainder of this section documents the evidence base for that change, including a comparison of imputation strategies, examples of suppressed values by industry, and the resulting impact on model predictions.


Strategies for Handling Suppressed Values

Three strategies for replacing suppressed QCEW values are show below to highlight the benefit of imputating the suppressed data. The first strategy is the previous process, which resulted in all suppressed values being zero. The second strategy uses the value from the prior period (either the prior quarter or the same month of the prior year) as a simple, rule-based imputation. The third strategy uses a gradient boosting imputation model trained on non-suppressed observations to generate a predicted value for each suppressed cell.

Each strategy was evaluated by simulating suppression on known values in the QCEW data and comparing the predicted values to the true employment counts. Root Mean Squared Error (RMSE) was used as the primary evaluation metric, as it penalizes large errors — which are precisely the cases of greatest concern for the SAM industry share calculations.

The two bar plots below show the RMSE for each strategy applied to quarterly and monthly suppressed values, respectively. Lower RMSE indicates better performance.




Across both the quarterly and monthly evaluations, the gradient boosting imputation model substantially outperforms the other two strategies. Setting suppressed values to zero — the previous approach — produces the highest RMSE, reflecting the large errors introduced when a suppressed value would have been substantial. The prior-period value approach improves on the zero-imputation strategy but still falls short of the model-based imputation. These results provided the basis for adopting the gradient boosting imputation model for the PY 2026-2027 SAM.


Imputation Results

This section illustrates the imputation model’s output for selected states, ownership levels, and industry sectors. In each plot below, the blue line and points represent the observed (non-suppressed) monthly employment counts over time. Orange points highlight months where the employment value was suppressed by BLS. For those suppressed months, the imputed value generated by the gradient boosting model is used in place of the original value (i.e., zero).

The examples were selected to highlight cases where suppression is non-trivial — i.e., where using zero in place of the suppressed value would meaningfully shift the industry share calculation. These cases are most common in smaller states and in industry-ownership combinations with relatively low employment counts.

Examples by Industry Sector
Construction

Manufacturing

Education and Healthcare



Using Imputed QCEW Data in Developing the SAM

The imputation process is applied before industry shares are calculated. For each state, year, and month, any suppressed ownership-industry employment count is replaced with the value predicted by the gradient boosting imputation model. Once suppressed values have been replaced, the industry share calculation proceeds as described in the previous section: ownership-level counts are summed by industry, and each industry’s total is divided by the total employment count across all non-Unclassified industries.

This means that the imputation affects the SAM in two distinct ways. First, it affects the fit data used to estimate the model coefficients, since the industry shares in the fitting sample are now calculated using imputed values where suppression occurred. Second, it affects the prediction data — both Estimate0 and Estimate1 — since industry shares for the prediction period are also constructed using imputed values for any suppressed cells. The impact on predictions is expected to be larger, as the most recent QCEW data (used for predictions) is preliminary and carries a higher suppression rate than the historical data used for model fitting.

Impact on Model Estimates

To evaluate the effect of imputation on model fit, two versions of the SAM fitting data were prepared: one using the imputed QCEW data and one using the original, non-imputed QCEW data. The SAM was run on both datasets for all program and indicator combinations, holding all other model inputs constant. This allows for a direct comparison of model fit between the two approaches, where any difference in performance can be attributed solely to the treatment of suppressed QCEW values.

The table below shows the RMSE for each program-indicator model under both the original and imputed approaches. RMSE measures the average magnitude of the difference between the model’s fitted values and the observed outcomes — a lower RMSE indicates a better-fitting model

Model Statistics: Original vs. Imputed QCEW Data, for Each Program and Indicator Model
Indicator Program RMSE (Original) RMSE (Imputed)
Credential Attainment
Credential Attainment Adult 0.0487 0.0442
Credential Attainment Dislocated Worker 0.0822 0.0792
Credential Attainment Youth 0.0534 0.0515
Employment Rate 2nd Quarter after Exit
Employment Rate 2nd Quarter after Exit Adult 0.0236 0.0226
Employment Rate 2nd Quarter after Exit Dislocated Worker 0.0340 0.0324
Employment Rate 2nd Quarter after Exit Wagner-Peyser 0.0303 0.0283
Employment Rate 2nd Quarter after Exit Youth 0.0378 0.0325
Employment Rate 4th Quarter after Exit
Employment Rate 4th Quarter after Exit Adult 0.0282 0.0248
Employment Rate 4th Quarter after Exit Dislocated Worker 0.0451 0.0431
Employment Rate 4th Quarter after Exit Wagner-Peyser 0.0217 0.0207
Employment Rate 4th Quarter after Exit Youth 0.0429 0.0321
Measurable Skill Gains
Measurable Skill Gains Adult 0.1078 0.0566
Measurable Skill Gains Dislocated Worker 0.1259 0.0742
Measurable Skill Gains Youth 0.1709 0.0629
Median Earnings 2nd Quarter after Exit
Median Earnings 2nd Quarter after Exit Adult 843.0648 778.6295
Median Earnings 2nd Quarter after Exit Dislocated Worker 1161.6482 955.7796
Median Earnings 2nd Quarter after Exit Wagner-Peyser 468.4394 345.8279
Median Earnings 2nd Quarter after Exit Youth 663.8769 487.6530
Note:
Green cells indicate that using imputed QCEW data improves the metric (lower RMSE) relative to the model without imputation. Red cells indicate the metric did not improve. RMSE for Median Earnings Q2 is expressed in dollars; all other RMSE values are in percentage points. Statistics are based on PY 2024 test data (PY 2024 annual outcomes).


Impact of Imputed QCEW Data on Model Predictions

This section presents the impact that using imputed QCEW data has on SAM predictions by comparing model predictions generated with and without imputed data against actual PY 2024 state performance outcomes.

Two key factors determine the magnitude of this impact. First, predictions are generated from a single observation of the QCEW variables for each state. If that observation contains a suppressed value that would otherwise be large, replacing it with zero can substantially distort the relevant industry share and, in turn, the model’s prediction for that state. Second, the prediction data — particularly for the most recent quarters — is drawn from preliminary QCEW releases, which have higher suppression rates than finalized data. As a result, the imputation has a more pronounced effect on predictions than on model coefficients.

Overall Result of Using a Model with Imputation

The bar chart below summarizes the prediction results across all program-indicator combinations. Each observation represents a state-level prediction for a given WIOA program and performance indicator in PY 2024. Predictions are classified into three categories:

  • Using a model with imputation resulted in predictions closer to actual performance — the absolute prediction error is smaller when using imputed data.
  • Not a significant difference between predictions — defined as a difference of less than $100 for Median Earnings 2nd Quarter after Exit, and less than 1 percentage point for all other indicators.
  • Using a model with imputation did not result in predictions closer to actual performance — the absolute prediction error is larger when using imputed data.

As the chart shows, in approximately 35% of cases the model using imputed data produced predictions closer to actual state performance outcomes. In approximately 59% of cases, there was no meaningful difference in predictions between the two approaches — that is, suppressed values in those observations were small enough that the choice of imputation strategy had negligible effect. In only about 6% of cases did the imputed model produce a less accurate prediction. Overall, these results demonstrate that the imputation process improves or preserves prediction accuracy in nearly all instances, with the greatest gains concentrated in cases where suppressed values are large.



Prediction Plots by Program

The plots below provide a more detailed view of the prediction comparison at the state level, organized by WIOA program. For each program and performance indicator, predictions from the model using imputed data and predictions from the model not using imputed data are plotted against the actual PY 2024 state performance outcomes.

The y-axis is the outcome value for the performance indicator — a percentage for all indicators except Median Earnings 2nd Quarter after Exit, which is expressed in dollars. The x-axis represents the 52 state observations, sorted from smallest to largest actual outcome value for that indicator. Sorting by actual outcome facilitates visual comparison of how well each model tracks true performance across the distribution of states.

Plot elements:

  • Blue — actual state performance in PY 2024.
  • Orange — model prediction using imputed QCEW data.
  • Purple — model prediction without imputed QCEW data.
  • Gray — observations where predictions from both approaches are not significantly different (as defined above). In these cases, both the orange and purple points are shown in gray, since the choice of imputation strategy does not meaningfully affect the prediction.

Vertical lines connect each prediction point to the actual outcome value, making it easier to identify which predictions are farthest from true performance.

Each tab contains plots for all performance indicators for that WIOA program.

Adult

WIOA Adult

Dislocated Worker

WIOA Dislocated Worker

Youth

WIOA Youth

Wagner-Peyser

Wagner-Peyser

Across all program indicators, the plots consistently show that predictions generated using imputed QCEW data more closely track actual state performance outcomes. The most visible pattern is in the outlier predictions — cases where the non-imputed model’s prediction falls far from the actual outcome. In nearly every such instance, the imputed model’s prediction is substantially closer to the truth. This pattern is most pronounced for the Measurable Skill Gains indicator, which is the performance measure most sensitive to suppressed QCEW values, likely due to the combination of industries and ownership types most relevant to workforce training programs.

Taken together, the bar chart and prediction plots provide strong evidence that the imputation process improves the accuracy and reliability of SAM predictions, particularly for states and time periods where suppressed QCEW values would have been large under the previous zero-replacement approach.







Prediction Plots

The plots below show performance of the selected PY 2026-2027 models in predicting actual PY 2024 performance. Note: these predictions are different than the Pre-Program Year Performance Estimates for the PY 2026-2027 models and are a test of model reliability using all data aligned with PY 2024.

Adult

Dislocated Worker

Youth

Wagner-Peyser

Final Estimates

This section has tables that provide the coefficients for each variable in each program indicator model. The last tab (All Model Estimates) has the complete data for all models and the table can be sorted, searched, and exported.

Adult

Term Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit Credential Attainment Measurable Skill Gains
female -0.023 -2398 0.021 -0.084 -0.013
age2544 -0.053 -608 -0.119 -0.057 -0.189
age4554 -0.046 -287 -0.206 -0.045 -0.196
age5559 -0.254 -2010 -0.314 -0.154 -0.162
age60 -0.364 -623 -0.539 -0.259 -0.082
hispanic -0.149 552 -0.112 0.015 -0.095
raceasian 0.271 5678 0.109 -0.024 -0.179
raceblack -0.128 -1084 -0.119 -0.128 0.034
racehpi -0.183 2740 -0.012 -0.295 0.313
raceai -0.078 -5220 -0.140 -0.305 -0.207
racemulti -0.215 1635 -0.073 -0.310 -0.144
hsgrad 0.148 392 0.166 0.069 0.125
collegedropout 0.072 -1618 0.035 -0.016 0.099
certotherps -0.052 -3496 0.144 0.139 0.154
associate 0.398 -1086 0.269 0.054 0.203
ba -0.015 3728 0.180 -0.179 -0.108
gradschool 0.327 9173 0.099 0.059 0.283
empentry 0.092 2103 0.058 0.149 0.006
edstatentry 0.027 1992 0.109 0.006 0.030
disabled -0.154 -3503 -0.144 -0.181 -0.057
veteran -0.233 -1516 -0.169 0.056 -0.566
englearner -0.091 -766 -0.005 -0.075 0.206
singleparent 0.033 2119 0.087 0.106 0.008
lowinc -0.017 -539 -0.053 0.059 0.018
homeless -0.104 -713 -0.209 -0.141
offender 0.019 1200 0.056 0.098 -0.109
dishomemaker -0.116 -4545 -0.283 -0.293
recwagesprior 0.144 -961 0.136 -0.013 0.068
longtermunemp -0.020 -978 -0.020 -0.063 0.131
uiclaimant 0.001 -44 0.009 0.061 -0.107
uiexhaustee 0.047 -1356 -0.090 -0.006
recotherasst -0.036 575 0.009 0.042 0.049
recssi -0.041 2653 -0.033 0.057 -0.182
rectanf -0.084 -2912 -0.074 -0.132 -0.073
daysinprog_morethanone 0.003 1327 -0.018 0.005
natresources -1.468 39477 -0.689 -3.776 3.829
construction -2.946 77101 -0.097 -4.929 15.214
manufacturing -0.218 110416 2.567 -6.620 5.593
tradeutil -0.911 73166 1.614 -3.949 11.117
information -2.589 20598 -1.505 3.971 10.472
financial -4.177 22853 -0.652 -9.197 10.110
business -0.543 110537 1.517 -3.923 10.744
edhealthcare -1.196 54159 1.273 -2.980 16.244
leisure -1.453 81458 1.037 -2.867 7.688
publicadmin -0.374 115168 3.222 -2.495 10.036
ur -1.353 -8812 -1.796 0.287 -3.184
wagesprior 3581
state_avg_wages 3713
daysenrolled -0.544
daysenrolled_under30 -0.270
AL 1.895 -70619 -0.577 4.801 -9.925
AK 1.920 -64531 -0.515 4.204 -10.028
AZ 2.039 -66855 -0.426 4.591 -10.176
AR 1.885 -69639 -0.585 4.716 -10.076
CA 1.965 -66906 -0.382 4.333 -10.070
CO 2.024 -66592 -0.377 4.490 -10.121
CT 1.979 -65049 -0.422 4.587 -10.306
DE 2.126 -64356 -0.407 4.738 -10.767
DC 1.705 -71587 -0.653 3.546 -9.195
FL 2.088 -66711 -0.351 4.582 -10.121
GA 1.965 -68532 -0.483 4.590 -10.160
HI 1.942 -69494 -0.468 4.157 -10.259
ID 1.950 -68616 -0.449 4.544 -10.055
IL 1.929 -67106 -0.480 4.620 -10.076
IN 1.846 -70686 -0.604 4.808 -9.726
IA 1.958 -67483 -0.530 4.865 -9.985
KS 1.902 -69409 -0.564 4.677 -10.058
KY 1.811 -69869 -0.656 4.630 -9.983
LA 1.979 -66866 -0.448 4.581 -10.341
ME 1.849 -67532 -0.586 4.488 -10.515
MD 1.951 -69506 -0.472 4.348 -10.519
MA 1.999 -65171 -0.380 4.425 -10.694
MI 1.902 -69753 -0.517 4.849 -9.933
MN 1.919 -66180 -0.519 4.743 -10.196
MS 1.910 -70035 -0.554 4.665 -10.218
MO 1.923 -67809 -0.503 4.557 -10.278
MT 1.921 -65489 -0.508 4.131 -10.199
NE 1.962 -67418 -0.486 4.641 -10.199
NV 2.068 -69300 -0.334 4.495 -9.617
NH 1.909 -67091 -0.507 4.656 -10.021
NJ 1.861 -67199 -0.536 4.446 -10.289
NM 1.962 -66269 -0.427 4.230 -10.340
NY 1.970 -63562 -0.411 4.292 -10.579
NC 1.919 -69362 -0.537 4.568 -10.139
ND 2.009 -62820 -0.399 4.510 -10.194
OH 1.916 -68503 -0.506 4.741 -10.034
OK 1.881 -68974 -0.577 4.536 -10.060
OR 1.859 -67616 -0.472 4.543 -9.971
PA 1.919 -67000 -0.484 4.551 -10.311
RI 2.015 -67031 -0.449 4.554 -10.387
SC 1.900 -70846 -0.546 4.648 -9.944
SD 1.963 -66474 -0.504 4.639 -10.223
TN 1.896 -69944 -0.555 4.616 -9.986
TX 2.035 -66906 -0.395 4.553 -10.174
UT 1.971 -68479 -0.466 4.566 -10.391
VT 1.879 -66719 -0.610 4.437 -10.521
VA 1.917 -70392 -0.479 4.478 -10.200
WA 1.864 -66346 -0.435 4.326 -10.207
WV 1.821 -67147 -0.561 4.344 -10.555
WI 1.831 -70182 -0.627 4.745 -9.872
WY 2.039 -61186 -0.349 4.282 -9.919
PR 1.988 -77417 -0.680 4.498 -9.869

Dislocated Worker

Term Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit Credential Attainment Measurable Skill Gains
female -0.047 -1510 -0.023 -0.104 -0.030
age2544 -0.186 -231 -0.190 0.038 -0.089
age4554 -0.047 -1623 -0.261 0.181 -0.233
age5559 -0.237 -1400 -0.289 0.090 -0.332
age60 -0.155 -5911 -0.604 -0.060 -0.201
hispanic -0.015 660 -0.023 -0.091 0.138
raceasian 0.054 -1659 0.066 -0.224 -0.012
raceblack -0.094 -190 -0.114 -0.132 0.071
racehpi 0.044 -8216 -0.138 -0.594 0.128
raceai 0.025 -5816 -0.031 -0.081 0.265
racemulti -0.017 -3244 -0.003 -0.073 -0.056
hsgrad 0.074 1129 0.028 0.127 0.255
collegedropout 0.059 522 -0.007 0.045 0.275
certotherps -0.007 -5211 -0.135 0.038 -0.042
associate 0.072 1065 0.062 0.147 0.311
ba 0.031 3260 0.058 -0.089 -0.068
gradschool 0.094 4504 0.133 0.120 0.219
empentry -0.022 346 -0.020 0.134 0.014
edstatentry -0.031 1635 0.068 0.011 0.005
disabled -0.007 -3297 0.078 -0.154 -0.061
veteran -0.116 745 -0.049 0.028 0.269
englearner -0.164 -1773 -0.200 -0.031 0.202
singleparent 0.143 -1760 0.069 -0.129 -0.039
lowinc -0.009 -416 -0.014 0.041 -0.064
homeless 0.467 580 0.268 -0.214
offender -0.040 1187 -0.100 0.010 0.009
dishomemaker -0.067 -1732 -0.044 0.131
recwagesprior 0.063 -1909 0.136 -0.178 0.064
longtermunemp -0.081 1136 -0.014 -0.030 -0.050
uiclaimant 0.033 563 -0.009 0.034 0.006
uiexhaustee 0.098 248 -0.092 -0.086
recotherasst 0.124 802 -0.010 -0.231 -0.126
recssi -0.136 -560 0.061 0.247 0.042
rectanf -0.237 973 -0.293 -0.040 0.074
daysinprog_morethanone 0.011 1172 0.028 0.011
natresources -0.016 -244836 1.407 -13.265 -12.193
construction -5.091 -132839 -1.723 -4.870 1.847
manufacturing -1.232 -138269 0.339 -8.231 -7.907
tradeutil -1.053 -135571 -0.121 -8.370 1.123
information 1.707 -152629 -1.692 -11.363 -5.297
financial -1.823 -172670 -0.671 -13.732 -4.275
business -1.263 -104765 1.575 -8.296 -1.791
edhealthcare -0.605 -163203 0.712 -5.863 1.937
leisure -1.164 -103423 0.571 -9.605 -6.712
publicadmin -2.235 -68626 1.399 -2.750 -2.843
ur -1.715 13139 -1.610 -0.824 -4.120
wagesprior 9668
state_avg_wages 4089
daysenrolled -0.218
daysenrolled_under30 -0.162
AL 2.235 136977 0.640 8.495 2.632
AK 2.289 140495 0.538 7.858 2.392
AZ 2.232 137690 0.626 8.607 2.297
AR 2.176 139431 0.650 8.527 2.477
CA 2.026 138346 0.556 8.553 2.459
CO 2.137 137429 0.606 8.558 2.429
CT 2.070 140744 0.660 8.432 2.266
DE 2.179 140971 0.619 8.652 1.910
DC 2.195 109872 0.219 6.753 2.820
FL 2.305 135581 0.692 8.638 2.262
GA 2.198 137377 0.659 8.585 2.281
HI 2.137 134738 0.588 8.254 2.216
ID 2.223 140030 0.588 8.369 2.466
IL 2.117 137714 0.644 8.559 2.325
IN 2.114 138438 0.619 8.399 2.775
IA 2.175 141534 0.718 8.506 2.623
KS 2.168 140668 0.618 8.491 2.462
KY 2.185 138324 0.605 8.372 2.665
LA 2.192 139074 0.624 8.495 2.336
ME 2.085 140054 0.598 8.273 2.105
MD 2.271 132824 0.574 7.938 1.981
MA 2.052 139882 0.577 8.385 1.968
MI 2.224 137699 0.677 8.593 2.647
MN 2.134 142100 0.624 8.524 2.478
MS 2.132 138179 0.616 8.422 2.419
MO 2.110 138474 0.618 8.462 2.311
MT 2.104 136638 0.535 8.024 2.220
NE 2.240 140093 0.697 8.392 2.410
NV 2.316 131878 0.703 8.829 3.011
NH 2.109 138622 0.655 8.435 2.313
NJ 1.977 136293 0.502 8.389 2.034
NM 2.108 138681 0.456 8.197 2.303
NY 1.917 137224 0.495 8.224 1.893
NC 2.152 135792 0.567 8.379 2.418
ND 2.171 147044 0.683 8.589 2.545
OH 2.134 138747 0.625 8.502 2.510
OK 2.118 139020 0.501 8.496 2.561
OR 2.015 138744 0.502 8.386 2.503
PA 2.113 139359 0.641 8.373 2.169
RI 2.180 137138 0.665 8.336 2.217
SC 2.225 136107 0.669 8.471 2.626
SD 2.093 140846 0.603 8.360 2.476
TN 2.205 136299 0.670 8.452 2.432
TX 2.138 140466 0.639 8.705 2.270
UT 2.180 139435 0.684 8.491 2.138
VT 2.063 139858 0.548 8.186 2.217
VA 2.261 133050 0.602 8.273 2.211
WA 1.997 140754 0.571 8.585 2.372
WV 2.115 139715 0.566 8.150 2.048
WI 2.129 139968 0.634 8.438 2.818
WY 2.236 146553 0.600 8.313 2.812
PR 2.138 128361 0.469 8.173 2.540

Youth

Term Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit Credential Attainment Measurable Skill Gains
female -0.103 -772 -0.152 -0.006 0.080
hispanic -0.064 -216 0.062 -0.061 -0.048
raceasian 0.004 -528 -0.257 0.068 -0.029
raceblack -0.074 -1220 -0.121 -0.063 -0.162
racehpi 0.275 -1816 -0.277 -0.119 0.039
raceai -0.142 -439 0.001 0.166 0.119
racemulti -0.072 -604 -0.012 0.275 0.029
hsgrad 0.142 844 0.155 0.188 0.053
collegedropout 0.104 -321 0.250 -0.014 -0.620
certotherps -0.391 2784 -0.226 -0.235 0.051
empentry 0.176 1626 0.126 0.000 0.135
edstatentry 0.142 -333 0.158 0.181 0.126
disabled -0.076 211 -0.081 0.035
englearner -0.077 -542 -0.170 0.071
lowinc -0.023 -496 -0.067 -0.047
homeless 0.049 -1297 -0.077 -0.061
offender -0.057 94 0.008 -0.037
longtermunemp -0.043 215 -0.029 -0.170 -0.005
uiclaimant 0.190 -20 0.089 0.072
recotherasst 0.106 -130 -0.005 0.054 -0.202
recssi -0.042 -3699 0.000 0.149 0.252
rectanf 0.133 -212 0.182 -0.126 0.101
daysinprog_morethanone 0.043 1157 0.031 0.007
natresources -3.180 -57744 2.603 -3.928 9.029
construction -2.748 20518 3.622 5.033 37.358
manufacturing -1.852 2577 4.653 2.559 16.590
tradeutil -0.011 -21182 6.890 -1.130 17.028
information -2.219 7386 -2.950 -17.099 3.594
financial -4.889 13072 2.603 6.178 26.075
business -1.188 -11610 4.863 -1.002 18.400
edhealthcare -3.206 -26421 2.218 0.787 23.064
leisure -2.463 798 4.659 1.294 18.888
publicadmin 1.717 60653 7.071 -1.315 21.129
ur -1.271 -5312 -1.025 -0.291 -3.631
state_avg_wages 3113
daysenrolled -0.355
daysenrolled_under30 -0.018
age1415 -0.106 -1802 -0.158 -0.553 0.142
age1617 -0.173 -1888 -0.077 -0.005 0.039
age1819 -0.129 -1602 -0.140 0.036 -0.015
age2021 -0.334 321 -0.178 -0.033 -0.045
associateorba 0.133 1956 -0.237 -0.375 -0.025
yparent 0.041 197 -0.076 -0.045 0.007
basiclitdeficient 0.039 229 0.029 -0.026 -0.054
yfoster -0.037 -2422 0.103 -0.043
ynaa 0.021 846 0.031 0.036 0.116
AL 2.592 9499 -3.445 0.093 -18.996
AK 2.308 8555 -3.594 0.435 -18.949
AZ 2.705 11305 -3.384 0.135 -19.436
AR 2.648 11541 -3.403 0.100 -19.132
CA 2.714 12518 -3.178 0.630 -18.534
CO 2.658 10297 -3.272 0.426 -19.183
CT 2.862 12885 -3.049 0.276 -18.852
DE 2.832 9938 -3.154 0.015 -19.774
DC 1.769 -3334 -3.397 1.106 -17.422
FL 2.744 10656 -3.291 0.389 -19.214
GA 2.632 10955 -3.312 0.559 -18.778
HI 2.523 10307 -3.345 0.196 -19.265
ID 2.678 10978 -3.336 0.068 -19.374
IL 2.655 11990 -3.302 0.265 -18.642
IN 2.681 11152 -3.354 0.003 -18.724
IA 2.751 10521 -3.272 -0.206 -19.261
KS 2.652 10088 -3.366 0.139 -19.182
KY 2.546 11225 -3.520 0.057 -18.836
LA 2.693 12173 -3.261 0.153 -19.459
ME 2.648 11307 -3.288 -0.020 -19.517
MD 2.498 8380 -3.347 0.380 -19.432
MA 2.803 12768 -3.061 0.376 -19.257
MI 2.698 12149 -3.272 0.178 -18.770
MN 2.701 12792 -3.200 0.062 -19.135
MS 2.736 11042 -3.309 0.278 -18.893
MO 2.760 11830 -3.203 0.159 -19.191
MT 2.517 9527 -3.501 -0.153 -19.588
NE 2.725 10756 -3.270 -0.023 -19.424
NV 2.680 9971 -3.531 0.015 -19.253
NH 2.624 11711 -3.298 0.307 -18.852
NJ 2.575 11966 -3.393 0.356 -18.636
NM 2.657 12017 -3.332 0.222 -19.446
NY 2.758 11357 -3.009 0.391 -19.087
NC 2.645 10305 -3.335 0.120 -19.104
ND 2.739 14661 -3.269 0.138 -19.102
OH 2.696 11639 -3.283 0.053 -18.922
OK 2.584 10153 -3.425 0.253 -19.023
OR 2.589 12577 -3.360 0.239 -18.921
PA 2.701 12180 -3.231 0.182 -18.867
RI 2.808 10672 -3.160 0.049 -19.368
SC 2.653 10149 -3.371 0.158 -18.906
SD 2.721 9642 -3.303 -0.200 -19.627
TN 2.616 11258 -3.391 0.173 -18.929
TX 2.726 12605 -3.291 0.134 -19.141
UT 2.714 9876 -3.321 0.279 -19.565
VT 2.675 11474 -3.270 0.021 -19.536
VA 2.584 9690 -3.280 0.386 -18.966
WA 2.577 11894 -3.167 0.723 -18.642
WV 2.552 11422 -3.371 0.335 -19.276
WI 2.681 10680 -3.296 -0.029 -18.866
WY 2.718 11692 -3.303 0.384 -19.223
PR 2.245 5668 -3.883 0.100 -18.815

Wagner-Peyser

Term Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit
female 0.006 -5654 0.125
age2544 -0.475 771 -0.407
age4554 -0.334 -271 -0.451
age5559 -0.373 -1447 -0.506
age60 -0.430 762 -0.397
hispanic 0.044 398 0.052
raceasian 0.020 -313 0.101
raceblack -0.111 -4210 -0.086
racehpi -1.186 -5540 -0.233
raceai -0.001 -2102 -0.050
racemulti -0.176 1003 -0.200
hsgrad 0.095 -5 0.021
collegedropout -0.118 73 -0.019
certotherps 0.283 1923 0.682
associate 0.004 -649 0.027
ba -0.124 2234 -0.065
gradschool -0.469 704 0.197
empentry 0.089 213 0.115
edstatentry -0.072 1363 -0.056
disabled 0.001 -1757 -0.021
veteran 0.086 67 0.028
englearner -0.104 -392 -0.066
singleparent 0.059 23 0.214
lowinc 0.009 -8 0.008
homeless -0.316 -5276 -0.303
offender 0.104 1853 -0.064
dishomemaker -0.559 -5190 -0.531
recwagesprior 0.439 485 0.227
longtermunemp -0.056 225 -0.112
uiclaimant -0.029 -139 -0.008
uiexhaustee 0.039 1737 0.029
recotherasst 0.414 5509 0.085
recssi -0.553 -5244 0.444
rectanf -0.149 -5229 -0.368
daysinprog_morethanone 0.045 1433 0.072
natresources 1.984 79450 -0.682
construction 1.218 101852 0.792
manufacturing -0.052 85969 0.285
tradeutil 2.171 106267 1.765
information 0.598 80561 0.190
financial -1.772 43394 -1.093
business 3.813 122729 2.124
edhealthcare 0.672 51508 0.885
leisure 2.969 90959 1.239
publicadmin 2.847 107555 1.998
ur -1.520 -8050 -2.091
wagesprior 4703
state_avg_wages 2774
AL -0.829 -78451 -0.289
AK -0.980 -77741 -0.299
AZ -0.879 -78783 -0.291
AR -0.769 -77795 -0.240
CA -0.968 -78027 -0.322
CO -0.980 -79052 -0.305
CT -0.663 -73063 -0.174
DE -0.721 -73410 -0.193
DC -1.287 -76989 -0.452
FL -0.960 -78905 -0.314
GA -0.869 -78503 -0.260
HI -0.877 -77501 -0.313
ID -0.878 -79457 -0.236
IL -0.823 -76963 -0.204
IN -0.664 -76909 -0.161
IA -0.571 -75886 -0.104
KS -0.757 -77910 -0.254
KY -0.838 -78351 -0.261
LA -0.881 -76637 -0.277
ME -0.765 -76395 -0.248
MD -1.019 -78359 -0.362
MA -0.755 -74471 -0.221
MI -0.747 -77585 -0.180
MN -0.690 -76174 -0.222
MS -0.709 -76240 -0.158
MO -0.782 -77130 -0.238
MT -0.950 -77383 -0.298
NE -0.730 -77430 -0.165
NV -1.112 -81853 -0.255
NH -0.795 -77346 -0.260
NJ -0.950 -77407 -0.376
NM -1.090 -78857 -0.356
NY -0.690 -72611 -0.192
NC -0.827 -77984 -0.237
ND -0.787 -76488 -0.177
OH -0.699 -76836 -0.162
OK -1.002 -79768 -0.318
OR -0.837 -77854 -0.208
PA -0.760 -75669 -0.226
RI -0.726 -74687 -0.194
SC -0.851 -78022 -0.231
SD -0.699 -75725 -0.207
TN -0.866 -79101 -0.262
TX -0.887 -78211 -0.273
UT -0.852 -79394 -0.259
VT -0.788 -75701 -0.317
VA -1.001 -79321 -0.300
WA -0.815 -77701 -0.189
WV -0.930 -78109 -0.256
WI -0.637 -76886 -0.163
WY -0.981 -77667 -0.197
PR -1.080 -85996 -0.480

All Model Estimates

Final Predictions

The tables below show the predicted outcomes (Pre-Program Year Performance Estimates) in PY 2026-2027 for each program indicator. The predictions are calculated by applying the model estimates to the most recent reported PY data on each state’s participant characteristics for each program indicator (i.e., PY 2024) and the most recent economic conditions data for states (i.e., time period 7/1/2024 - 6/30/2025).

Adult

State Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit Credential Attainment Measurable Skill Gains
AK 0.845 13883 0.831 0.726 0.734
AL 0.880 11005 0.864 0.800 0.752
AR 0.785 8241 0.756 0.760 0.802
AZ 0.700 9492 0.667 0.757 0.769
CA 0.687 8758 0.665 0.730 0.805
CO 0.702 10158 0.690 0.822 0.707
CT 0.746 8662 0.744 0.731 0.830
DC 0.686 10488 0.687 0.534 0.866
DE 0.807 9152 0.755 0.677 0.444
FL 0.846 10740 0.831 0.755 0.801
GA 0.826 9032 0.815 0.744 0.737
HI 0.722 9119 0.730 0.514 0.530
IA 0.776 8599 0.763 0.761 0.744
ID 0.702 8491 0.696 0.740 0.785
IL 0.805 10037 0.791 0.752 0.701
IN 0.777 8776 0.766 0.765 0.838
KS 0.709 8868 0.698 0.791 0.689
KY 0.752 8558 0.750 0.721 0.755
LA 0.715 8745 0.717 0.768 0.821
MA 0.709 8651 0.719 0.674 0.521
MD 0.803 9842 0.803 0.675 0.771
ME 0.678 8843 0.665 0.667 0.595
MI 0.826 9552 0.790 0.881 0.690
MN 0.759 10141 0.746 0.794 0.751
MO 0.734 8795 0.728 0.665 0.715
MS 0.886 9021 0.895 0.768 0.766
MT 0.691 10794 0.701 0.506 0.656
NC 0.781 8669 0.782 0.670 0.733
ND 0.790 10716 0.779 0.684 0.707
NE 0.764 9119 0.758 0.715 0.663
NH 0.855 12011 0.844 0.822 0.820
NJ 0.668 7717 0.650 0.676 0.744
NM 0.792 11089 0.792 0.689 0.794
NV 0.736 8954 0.724 0.803 0.809
NY 0.627 8106 0.663 0.618 0.824
OH 0.769 9086 0.777 0.796 0.774
OR 0.689 9598 0.681 0.760 0.647
PA 0.744 7898 0.739 0.716 0.759
PR 0.652 4727 0.639 0.594 0.705
RI 0.816 7802 0.796 0.747 0.709
SC 0.774 8488 0.751 0.701 0.725
SD 0.668 5995 0.681 0.654 0.679
TN 0.815 8759 0.797 0.714 0.776
TX 0.771 9023 0.748 0.712 0.709
UT 0.728 8751 0.724 0.698 0.549
VA 0.818 9204 0.800 0.746 0.834
VT 0.706 9476 0.661 0.678 0.664
WA 0.671 12131 0.689 0.750 0.527
WI 0.748 8358 0.751 0.701 0.680
WV 0.776 8659 0.773 0.773 0.639
WY 0.762 12664 0.704 0.719 0.704
OK 0.726 8448 0.710 0.744 0.788

Dislocated Worker

State Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit Credential Attainment Measurable Skill Gains
AK 0.902 15830 0.901 0.630 0.792
AL 0.833 10721 0.838 0.880 0.765
AR 0.799 9865 0.823 0.764 0.842
AZ 0.768 11125 0.712 0.787 0.749
CA 0.730 11169 0.718 0.814 0.742
CO 0.720 13065 0.711 0.804 0.667
CT 0.805 10983 0.808 0.767 0.801
DC 0.716 11683 0.687 0.451 0.743
DE 0.782 11151 0.779 0.754 0.442
FL 0.841 11842 0.841 0.815 0.722
GA 0.821 11729 0.826 0.798 0.663
HI 0.786 11223 0.761 0.576 0.512
IA 0.853 10701 0.858 0.753 0.769
ID 0.771 10500 0.754 0.658 0.676
IL 0.809 12105 0.799 0.788 0.736
IN 0.756 10190 0.755 0.773 0.843
KS 0.801 14047 0.844 0.901 0.796
KY 0.839 11671 0.799 0.742 0.870
LA 0.706 10629 0.701 0.845 0.799
MA 0.784 13716 0.793 0.697 0.525
MD 0.817 11644 0.824 0.674 0.767
ME 0.760 11011 0.791 0.721 0.597
MI 0.878 11129 0.849 0.888 0.739
MN 0.829 14403 0.826 0.911 0.794
MO 0.752 9691 0.754 0.737 0.724
MS 0.788 8404 0.797 0.756 0.767
MT 0.702 12868 0.713 0.561 0.630
NC 0.722 9757 0.732 0.650 0.751
ND 0.814 13613 0.907 0.809 0.924
NE 0.875 11743 0.863 0.716 0.812
NH 0.812 13760 0.842 0.811 0.622
NJ 0.683 10869 0.696 0.708 0.804
NM 0.735 10574 0.741 0.673 0.803
NV 0.816 11671 0.801 0.791 0.811
NY 0.666 9055 0.694 0.603 0.751
OH 0.769 12121 0.802 0.838 0.771
OR 0.683 9949 0.688 0.761 0.657
PA 0.827 10562 0.819 0.783 0.751
PR 0.623 5147 0.652 0.648 0.651
RI 0.820 10950 0.862 0.766 0.650
SC 0.802 11011 0.805 0.767 0.743
SD 0.739 9502 0.745 0.755 0.877
TN 0.826 10160 0.815 0.708 0.739
TX 0.730 11625 0.768 0.818 0.677
UT 0.785 12754 0.797 0.735 0.471
VA 0.833 10958 0.830 0.746 0.785
VT 0.766 13273 0.779 0.751 0.748
WA 0.713 13407 0.715 0.798 0.541
WI 0.811 10822 0.801 0.701 0.782
WV 0.833 11102 0.860 0.805 0.624
WY 0.818 16608 0.793 0.735 0.631
OK 0.759 11077 0.752 0.835 0.762

Youth

State Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit Credential Attainment Measurable Skill Gains
AK 0.560 4927 0.540 0.533 0.689
AL 0.700 4393 0.696 0.567 0.568
AR 0.731 4619 0.718 0.552 0.688
AZ 0.759 7317 0.746 0.677 0.778
CA 0.687 5244 0.668 0.637 0.741
CO 0.720 5866 0.708 0.722 0.558
CT 0.796 5321 0.782 0.798 0.815
DC 0.650 6248 0.604 0.547 0.676
DE 0.674 3984 0.734 0.705 0.695
FL 0.785 5356 0.790 0.813 0.738
GA 0.754 4800 0.750 0.747 0.632
HI 0.629 5549 0.630 0.628 0.537
IA 0.758 4895 0.747 0.525 0.526
ID 0.736 6489 0.771 0.559 0.777
IL 0.759 5741 0.770 0.706 0.676
IN 0.782 5202 0.784 0.686 0.811
KS 0.733 5470 0.736 0.657 0.526
KY 0.687 5980 0.683 0.683 0.754
LA 0.736 5393 0.758 0.711 0.660
MA 0.703 5127 0.703 0.632 0.508
MD 0.772 5945 0.734 0.659 0.660
ME 0.675 5194 0.680 0.558 0.505
MI 0.770 5963 0.759 0.694 0.591
MN 0.727 6269 0.744 0.591 0.600
MO 0.751 5442 0.746 0.611 0.650
MS 0.833 4382 0.849 0.804 0.734
MT 0.655 4954 0.603 0.390 0.600
NC 0.736 5318 0.733 0.563 0.663
ND 0.878 8327 0.800 0.653 0.710
NE 0.745 5443 0.741 0.543 0.529
NH 0.830 6703 0.835 0.702 0.771
NJ 0.616 3803 0.610 0.547 0.753
NM 0.665 5507 0.701 0.486 0.536
NV 0.683 6005 0.658 0.511 0.611
NY 0.644 4075 0.677 0.585 0.677
OH 0.734 4786 0.717 0.603 0.655
OR 0.602 5607 0.570 0.549 0.501
PA 0.696 4839 0.676 0.624 0.730
PR 0.610 3679 0.600 0.443 0.776
RI 0.705 4404 0.709 0.615 0.461
SC 0.785 5532 0.761 0.660 0.697
SD 0.730 4252 0.752 0.490 0.521
TN 0.787 6052 0.794 0.693 0.705
TX 0.751 5818 0.745 0.596 0.721
UT 0.770 6255 0.743 0.637 0.611
VA 0.794 5630 0.774 0.687 0.773
VT 0.731 5751 0.688 0.541 0.486
WA 0.626 5983 0.624 0.478 0.409
WI 0.774 5510 0.758 0.577 0.618
WV 0.670 5024 0.642 0.694 0.541
WY 0.729 4960 0.715 0.571 0.693
OK 0.745 6858 0.762 0.713 0.828

Wagner-Peyser

State Employment Rate 2nd Quarter after Exit Median Earnings 2nd Quarter after Exit Employment Rate 4th Quarter after Exit
AK 0.690 10221 0.664
AL 0.714 7180 0.732
AR 0.697 7801 0.690
AZ 0.619 8516 0.575
CA 0.578 9117 0.611
CO 0.632 9159 0.606
CT 0.636 8399 0.672
DC 0.557 8469 0.564
DE 0.668 8183 0.657
FL 0.658 8200 0.661
GA 0.656 7791 0.689
HI 0.615 10371 0.643
IA 0.727 9455 0.717
ID 0.697 9107 0.694
IL 0.644 8658 0.687
IN 0.712 8980 0.715
KS 0.675 8002 0.668
KY 0.603 8024 0.598
LA 0.622 7085 0.640
MA 0.621 10749 0.659
MD 0.626 9102 0.664
ME 0.590 8075 0.589
MI 0.711 8893 0.684
MN 0.616 10327 0.649
MO 0.656 7922 0.680
MS 0.782 6767 0.785
MT 0.498 8816 0.686
NC 0.692 8069 0.708
ND 0.624 8610 0.631
NE 0.680 8846 0.709
NH 0.735 11552 0.710
NJ 0.555 8467 0.576
NM 0.652 8362 0.639
NV 0.711 9193 0.697
NY 0.673 8939 0.699
OH 0.647 10726 0.642
OR 0.651 9159 0.671
PA 0.655 7736 0.692
PR 0.531 4353 0.525
RI 0.658 8938 0.684
SC 0.657 7659 0.657
SD 0.694 6692 0.685
TN 0.664 7943 0.661
TX 0.666 7958 0.679
UT 0.637 8852 0.628
VA 0.696 8340 0.696
VT 0.660 8780 0.607
WA 0.665 10843 0.676
WI 0.679 8849 0.662
WV 0.612 6898 0.636
WY 0.658 6875 0.662
OK 0.614 7083 0.624

All Predictions

Model Specification Data

The tabs below have information on which variables are included in the SAM for each program indicator model and about now normalization was applied to select variables.

Full Variable Table

This table is the full list of all the variables included in all the PY 2026-2027 models. An “x” indicates that the variable is included in that particular indicator model for that WIOA program.

Adult and Dislocated Worker
Youth
Wagner-Peyser
Full Variable Name ERQ2 ERQ4 MEQ2 CRED MSG Youth ERQ2 Youth ERQ4 Youth MEQ2 Youth CRED Youth MSG WP ERQ2 WP ERQ4 WP MEQ2
Female x x x x x x x x x x x x x
Age 14 to 15 x x x x x
Age 16 to 17 x x x x x
Age 18 to 19 x x x x x
Age 20 to 21 x x x x x
Age 25 to 44 x x x x x x x x
Age 45 to 54 x x x x x x x x
Age 55 to 59 x x x x x x x x
Age 60 or more x x x x x x x x
Hispanic Ethnicity x x x x x x x x x x x x x
Race: Asian x x x x x x x x x x x x x
Race: Black x x x x x x x x x x x x x
Race: Hawaiian or Pacific Islander x x x x x x x x x x x x x
Race: American Indian x x x x x x x x x x x x x
Race: Multiple x x x x x x x x x x x x x
Highest Grade Completed: High School Equivalency x x x x x x x x x x x x x
Highest Grade Completed: Some College x x x x x x x x x x x x x
Highest Grade Completed: Certificate or Other Post-Secondary Degree x x x x x x x x x x x x x
Highest Grade Completed: Associate Degree x x x x x x x x
Highest Grade Completed: Bachelor Degree x x x x x x x x
Highest Grade Completed: Associate or Bachelor Degree x x x x x
Highest Grade Completed: Graduate Degree x x x x x x x x
Employed at Program Entry x x x x x x x x x x x x x
In School at Program Entry x x x x x x x x x x x x x
Individual with a Disability x x x x x x x x x x x x
Veteran x x x x x x x x
Limited English Proficiency x x x x x x x x x x x x
Single Parent x x x x x x x x
Low Income x x x x x x x x x x x x
Homeless x x x x x x x x x x x
Individual who was Incarcerated x x x x x x x x x x x x
Displaced Homemaker x x x x x x x
Foster Care Youth x x x x
Youth Parent or Pregnant Youth x x x x x
Skills/Literacy Deficient at Program Entry x x x x x
Received Wages Prior to Participation x x x x x x x x
Wages Prior to Participation (Normalized) x x
Long-Term Unemployed at Program Entry x x x x x x x x x x x x x
UI Claimant x x x x x x x x x x x x
UI Exhaustee x x x x x x x
Received Other Public Assistance x x x x x x x x x x x x x
SSI or SSDI Recipient x x x x x x x x x x x x x
TANF Recipient x x x x x x x x x x x x x
Youth Needing Additional Assistance x x x x x
Median Days in Program (Normalized) x x x x x x x x x x x
Median Days Enrolled in Education or Training (Normalized) x x
Percent Enrolled in Education or Training Under 30 Days x x
Natural Resources Employment x x x x x x x x x x x x x
Construction Employment x x x x x x x x x x x x x
Manufacturing Employment x x x x x x x x x x x x x
Information Services Employment x x x x x x x x x x x x x
Financial Services Employment x x x x x x x x x x x x x
Professional and Business Services Employment x x x x x x x x x x x x x
Educational or Health Care Employment x x x x x x x x x x x x x
Leisure, Hospitality, or Entertainment Employment x x x x x x x x x x x x x
Public Administration x x x x x x x x x x x x x
Trade, Transportation, and Utilities x x x x x x x x x x x x x
Average Wages of Labor Force (Normalized within State) x x x
Unemployment Rate Not Seasonally Adjusted x x x x x x x x x x x x x

Normalization Data

All the variables used in the models that are not percentages (i.e., had a default value from 0 to 1) were normalized. This method was applied to all the data used to fit the models and for the prior values used to get the the predicted outcomes (Pre-Program Year Performance Estimate) for PY 2026-2027. Normalization will also be applied to the actual values when the models are applied for the performance assessments. The actual values will be normalized at the same scale by using the min max values in the table below.

A table with all the minimum and maximum values for the variables where normalization was applied are shown in the table below. The table can be exported as desired. Below the table is additional informaiton on how the normalization is applied to the data.

Normalization Method

The normalization method used is Min-Max normalization. This method converts all the values into a scale from 0 to 1 which aligns with most of the other model variables which are percentages and thus on the same scale.

Getting the Minimum and Maximum Values

The minimum and maximum values were obtained by getting the total data (i.e., data from PY 2019 - 2024) and then capturing the min-max values of the total data by program.

There is a slight variation in min-max values for the Average Wages of Labor Force variable. Unlike the other variables, which use the minimum and maximum for the variable from the total data, the Average Wages of Labor Force uses the min-max values within the state. In other words, the data is first grouped by state and then the min and max value for each state for Average Wages of Labor Force is used. This approach is taken because the value of the variable is to capture wage changes in the state rather than get the relative wages of a state compared to other states.

The data in the table can be used to convert raw values into normalized values or normalized values back to the raw values. For example, if you had a Wages Prior to Participation value of 6,500 for the Adult program and MEQ2 Indicator you could normalize that value to the scale that was used for the PYs 2026-2027 models. If you look at the data in the table below, that variable had a minimum value of 1,273 and a maximum value of 12,519. Applying the min-max formula using those values gives a normalized value of 0.465. Likewise, if you had a normalized value of 0.7 for the same program and indicator you could apply the formula to get the original raw value of 9,145.