TOPIC LIBRARY · PERFORMANCE + DATA
Source + trust status
Automated source check completed September 22, 2026. Source basis: WSD24-07 Performance Guidance, WSD24-16 Data Validation, TEGL 07-18, Change 1, and TEN 03-26. Check result: The performance indicators, exit and cohort concepts, credential and MSG concepts, median-earnings treatment, data-validation framework, and statistical-adjustment concepts remain supported by current guidance. Current employer indicator: Effectiveness in Serving Employers is defined as Retention with the Same Employer. Assessment caution: the indicators assessed for sanctions can vary by program and program year, so current assessment notices should be checked before describing a measure as assessed. This AI-assisted page does not receive human legal or compliance review. Official sources and applicable local policy, agreement, grant terms, and procedures control. Trust standard →
Workforce Wonkery Learning · Functional track
Performance + Data
WIOA performance starts long before a performance report is produced. Eligibility, service codes, case notes, exits, credentials, skill gains, wage data, and follow-up all become part of the numbers that boards, states, and the federal government use to judge results.
30-second takeaway
- Not everyone who touches the workforce system becomes a participant. Reportable individuals and participants are treated differently for performance.
- Many WIOA measures are based on what happens after exit, while Measurable Skill Gains is an in-program measure.
- Median earnings is not an average, annual salary, or placement wage. It is the middle quarterly earnings value among participants included in that measure.
- A good outcome that is never entered or documented may not count.
- Local performance is judged against statistically adjusted goals, not simply by comparing raw outcomes with another Local Area.
Why this matters
Performance is not a separate back-office function. Front-line staff create the data, managers shape service strategy, MIS staff protect data quality, providers influence outcomes, and boards are accountable for the results.
How a person becomes a performance number
| Step | Stage | What it means |
|---|---|---|
| 1 | Reportable individual | The person engages with the system but has not yet met the definition of a participant. Self-service or information-only activity often stays at this level. |
| 2 | Participant | The person meets program requirements and receives a service that starts participation for that program. |
| 3 | Period of participation | The participant receives services during a defined period. A person can have more than one period of participation over time. |
| 4 | Exit | After 90 days without a clock-restarting service and with no future qualifying service scheduled, exit is applied retroactively to the last qualifying service date. |
| 5 | Cohort | Participants are grouped according to the rules for each measure, often by exit period or by whether they are in qualifying education or training during the program year. |
| 6 | Indicator | Wage records, supplemental wage information, education data, credentials, MSGs, and other reported data become the results used in WIOA performance. |
The six primary WIOA indicators
| Indicator | What it asks | Timing |
|---|---|---|
| Employment Rate, Q2 after exit | Was the participant in unsubsidized employment in the second quarter after exit? Youth can also count through qualifying education or training. | Exit-based rate |
| Employment Rate, Q4 after exit | Was employment, or qualifying Youth education/training, sustained into the fourth quarter after exit? | Exit-based rate |
| Median Earnings, Q2 after exit | What is the middle quarterly earnings value among participants in unsubsidized employment in the second quarter after exit? | Exit-based dollar amount |
| Credential Attainment | Did participants in qualifying education or training earn a recognized credential within the required time period? | Exit-based, with program-specific rules |
| Measurable Skill Gains | Did participants in qualifying education or training make at least one documented gain during the program year? | In-program annual rate |
| Effectiveness in Serving Employers | Retention with the Same Employer: among participants with wage data, how often are those employed in Q2 after exit also employed by the same employer in Q4? | Shared employer indicator; current federal measure is Retention with the Same Employer |
Three concepts people often misread
Median earnings, MSG, and the Statistical Adjustment Model
Median earnings · Why this number is not an average wage
Median means the middle, not the average. Put the quarterly earnings of the participants included in the measure in order from lowest to highest. The middle value is the median. If there is an even number of values, the median is the average of the two middle values.
Example: five employed participants have Q2 quarterly earnings of $5,000, $7,500, $9,000, $12,500, and $28,000. The median is $9,000. The $28,000 value does not pull the median upward the way it would pull up an arithmetic average.
| Do not read median earnings as… | Read it as… |
|---|---|
| Average earnings of every participant | The middle earnings value among participants included in the earnings measure |
| Annual salary | Quarterly earnings in Q2 after exit |
| Hourly wage at placement | Total covered earnings recorded for the applicable quarter |
| Employment rate | A separate measure of earnings among people who are employed in Q2 |
This creates an important interpretation issue: a Local Area can have a strong median-earnings result and a weak Q2 employment rate. If relatively few participants are employed but the employed participants earn well, those two measures tell different stories.
Board question: always look at Employment Q2 and Median Earnings together. One tells you how many are employed; the other tells you about the middle earnings level among those included in the earnings measure.
Measurable Skill Gains · Why one participant can matter in more than one program year
MSG is an in-program measure. It is intended to show whether participants in qualifying education or training are making documented progress while they are still participating.
The denominator is annual. A participant who remains in qualifying education or training across more than one program year can enter the MSG denominator in more than one program year. In general, staff should expect to document a new gain for each program year in which the participant remains in the applicable denominator.
Only one successful gain is needed to count for that program year. A participant may have several documented gains, but multiple gains do not make that participant count more than once in the numerator for the same program year.
| Scenario | What it means |
|---|---|
| Training starts before June 30 and continues after July 1 | The participant may be in the MSG denominator in two different program years, depending on the applicable program and activity rules. |
| One documented gain in the first program year | That can satisfy MSG for that program year, but it does not automatically satisfy the next program year. |
| Three gains in one program year | Excellent progress, but the participant still contributes one success to the MSG numerator for that program year. |
| No documented gain | The participant can remain in the denominator without a numerator success even if staff believe progress occurred. |
Five broad MSG types: educational functioning level gain; secondary diploma or recognized equivalent; qualifying transcript/report card progress; satisfactory progress toward a training milestone; and successful passage of an occupational exam or documented technical/occupational skill progress.
Most common mistake: treating MSG like a one-time credential measure. It is a recurring annual progress measure for participants who meet the applicable education/training conditions. Use WSD24-07’s detailed tables to determine who belongs in the denominator.
Statistical Adjustment Model · What it actually does
The SAM does not change the Local Area’s actual performance. It changes the benchmark used to judge that performance so the benchmark reflects the participant population actually served and the economic conditions in which the program operated.
Think of it as an evidence-based expectation model. Using historical WIOA data, DOL estimates how participant characteristics and labor-market conditions have been associated with each performance indicator. Those relationships become model coefficients. The model then asks: given this mix of participants and this economy, what level of performance would we statistically expect?
The basic sequence
1. Before the program year, the model estimates performance using prior participant data and recent economic conditions.
2. The state and DOL negotiate a performance goal using the model estimate plus other required factors.
3. After the program year, DOL looks at the actual participant mix and actual economic conditions experienced during the year.
4. The model calculates whether those actual conditions would statistically be expected to make the indicator easier or harder to achieve than originally anticipated.
5. That difference becomes an adjustment factor that is added to the negotiated goal.
6. The actual result is divided by the final adjusted goal to produce the performance score.
So what can move the expectation? The exact variables differ by program and indicator, but the PY 2026–27 models use participant characteristics such as:
| Participant characteristic | Examples used in current models |
|---|---|
| Age + demographics | Age groups, sex, race, and ethnicity. |
| Education | Highest grade or credential completed and, for relevant models, school status at program entry. |
| Employment history | Employment at program entry, prior wages, long-term unemployment, UI claimant status, and UI exhaustion. |
| Barriers + circumstances | Disability, limited English proficiency, single-parent status, low-income status, homelessness, incarceration history, and displaced-homemaker status. |
| Other program-entry characteristics | Veteran status, receipt of public assistance, SSI/SSDI, TANF, and other variables used when statistically supported for the specific program and indicator. |
| Program experience | Some current models also use measures such as median days in the program. |
The economic side matters too. Current models use the unemployment rate and the local/state industry mix, such as the shares of employment in construction, manufacturing, health care and education, leisure and hospitality, professional services, trade/transportation/utilities, public administration, and other major sectors. For Median Earnings, the model also incorporates labor-force wage information so the expected earnings benchmark can reflect changing wage levels.
A simple example of what the adjustment means
Suppose Local Area A negotiated a Q2 employment goal of 70%. When the year ends, the area’s actual participant population includes more people with characteristics that, in the model, have historically been associated with lower Q2 employment outcomes, and the unemployment rate is higher than the conditions used before the year.
The model does not increase the area’s actual employment result. Instead, it may conclude that the operating environment was harder than originally expected and lower the benchmark. If the final adjusted goal becomes 65% and the actual result is 68%, the Individual Indicator Score is:
68% actual ÷ 65% final adjusted goal = 104.6% performance score
The area did not “get credit” by changing its actual result. The model changed the expected benchmark because the participant mix and economy were different from what had been anticipated. A score above 100% means actual performance exceeded the final adjusted expectation.
The adjustment can also move the benchmark upward. If actual participant and economic conditions are statistically associated with stronger expected outcomes than the conditions assumed before the year, the final adjusted goal can increase.
| Number | What it means |
|---|---|
| Actual result | The employment, earnings, credential, or skill-gain outcome actually reported. The SAM does not alter it. |
| Negotiated goal | The performance level established before the year through the negotiation process. |
| Adjustment factor | The difference between what the model predicts using the pre-year conditions and what it predicts using the actual participant mix and economic conditions. |
| Final adjusted goal | Negotiated goal plus the adjustment factor. |
| Performance score | Actual result divided by the final adjusted goal. A score of 100% means the final adjusted goal was met. |
Bottom line: the SAM is an attempt to make performance accountability more comparable by adjusting the expectation for differences in participant mix and economic conditions that are partly outside a Local Area’s control. It does not erase weak outcomes, guarantee a downward adjustment, or make raw results irrelevant. Read the actual result, final adjusted goal, and performance score together. See WSD20-02 →
What front-line decisions change performance?
| Operational decision | Performance effect |
|---|---|
| Choosing an activity code | Some services start or restart the participation clock. A wrong code can create the wrong participation or exit date. |
| Entering a credential | A valid credential can improve credential attainment only if it meets the definition, is documented, and is entered correctly. |
| Recording an MSG | A participant can make real progress and still appear unsuccessful if the gain is not entered and supported by acceptable documentation. |
| Leaving services open | A service can delay exit if it is clock-restarting. An incorrect end date can shift the participant into a different cohort. |
| Co-enrolling across programs | Programs can sometimes share credit for outcomes, but data has to be entered in the correct programs and systems. |
| Collecting supplemental wage information | When wage-record matching does not capture employment, acceptable supplemental information can help document employment and earnings under federal rules. |
Performance deep dive
The definitions that change the numbers
Reportable individual vs. participant · Who enters performance?
A reportable individual may provide identifying information or use self-service or information-only services without becoming a participant. Reportable individuals are reported, but they are not included in participant performance calculations.
For Adult, DW, and several other programs, a person generally becomes a participant after meeting program requirements and receiving a service that starts participation. Youth participation has additional required components.
Clock-restarting services + exit · Why dates matter
For a participant, exit generally occurs after 90 consecutive calendar days without a clock-restarting service and with no future qualifying service scheduled. The exit date is applied retroactively to the last qualifying service date.
Follow-up, self-service, information-only activity, and supportive services do not by themselves delay or restart exit under the current performance rules. Use the activity-code policy rather than informal labels.
Common exit · One person, several programs
CalJOBS uses common exit across specified DOL-administered programs. A participant does not exit until they have gone 90 days without a clock-restarting service from any covered program and no future qualifying service is scheduled.
Why it matters: another program’s service can keep a participant from exiting even when your program stopped providing services.
Exclusionary exits · When someone is removed from certain measures
Certain exit reasons can exclude a participant from one or more performance measures, including specified situations involving incarceration or institutionalization, long-term medical treatment, death, qualifying military reserve activation, and a limited foster-care relocation situation for Youth.
Do not use exclusions as a performance strategy. The reason must actually meet the federal definition and be supported.
Credential attainment · What actually counts?
Recognized postsecondary credentials can include degrees, occupational licenses, apprenticeship completion, occupational certificates, and industry-recognized certifications that document technical or occupational skills.
General safety, work-readiness, or basic workplace certificates do not automatically meet the WIOA postsecondary credential definition. Ask whether the credential is recognized for technical or occupational skills and use the detailed WSD24-07 rules for the applicable program.
Wage records + supplemental wage information · Where employment data comes from
Employment and earnings are often captured through unemployment-insurance wage-record matching. When wage records do not capture the employment, federal rules allow specified supplemental wage information when collected and documented correctly.
Do not wait until reporting season. Build follow-up and wage-documentation practices into normal case management while customers are still reachable.
Data validation · Can the reported value be proven?
EDD conducts annual WIOA data validation to test whether data reported through CalJOBS is accurate, reliable, and supported by acceptable source documentation. Local Areas must be able to produce the documentation for sampled data elements.
Data quality is a compliance issue. If the source document does not support the data element, the problem can become a validation finding and can distort performance.
Show me how a small data error becomes a big performance problem
A participant completes occupational training and earns a recognized credential.
The provider sends proof to the case manager. The credential is uploaded to the file but never entered in the correct CalJOBS performance field. The annual report therefore treats the participant as having no credential.
The lesson: service delivery, documentation, and data entry are one chain. If one link breaks, the outcome may disappear from the performance report.
Before you trust a performance number
- Which program is being measured?
- Is this an actual result, negotiated goal, final adjusted goal, or performance score?
- Which exit cohort or program-year population is included?
- For a rate, what is the numerator and denominator?
- For median earnings, who is included and are you looking at a quarterly median rather than an average?
- For MSG, which participants entered the denominator this program year and which have a documented gain?
- Were exclusionary exits, credentials, co-enrollment outcomes, and supplemental wages handled correctly?
- Has the data passed source-document validation?
- Could service coding, exit timing, participant characteristics, or economic conditions explain the result?
Policy connections
| Topic | Policy brief | Why it matters |
|---|---|---|
| Performance definitions | WSD24-07 | Core California definitions for reportable individuals, participants, exit, credentials, MSG, wage outcomes, and the WIOA indicators. |
| Statistical adjustment + nonperformance | WSD20-02 | How actual outcomes, final SAM-adjusted goals, Individual Indicator Scores, and nonperformance thresholds fit together. |
| Activity codes | WSD24-05 | Which CalJOBS activities start or restart participation and how services are reported. |
| Participant reporting | WSD20-10 | Reporting timelines and data-entry responsibilities in CalJOBS. |
| Data validation | WSD24-16 | How EDD tests whether reported participant data is supported by source documentation. |
| Current goals | WSIN26-04 | The current state-level negotiated goals for PY 2026 and PY 2027. |
| Current scores | WSIN25-38 | How California reports Local Area performance scores after adjustment. |
| Federal assessment | TEN 03-26 | The current federal performance-accountability assessment context for PY 2025. |
What to read next
Need to see how case decisions become data? Customer Journey →
Need to see how the data gets tested? Monitoring + Compliance →
Need to manage provider outcomes? Contracts + Subrecipient Management →
