Appendix A: Description of the Data, Analysis Notes, and Results
Description of the Data
Data for this analysis were pulled out of the county human resource data management system, Workday. Information for all individuals employed by the county any time between July 1, 2021 and June 30, 2024 was included (FY 2022 – FY 2024). Data were pulled in two different reports. The first, the ‘snapshot’ data, is the employee demographics on the last day of each fiscal year (June 30). This is a snapshot of a specific point in time of our workforce. The second data source, the ‘actions’ dataset, looks at employee Human Resource (HR) actions that occur over the course of a given fiscal year. This data set includes all the actions that occurred throughout the 3-year period rather than on a particular day; it also includes information on all employees who were at the county at any time during the 3-year period, even if they had no HR actions. More detailed information about the data used in each report is outlined below:
- Demographics Report – For the analyses that show demographic changes over time (i.e., show data for the years: FY 2022, FY 2023, and FY 2024 as well as the reference years of FY 2020 and FY 2021 that are occasionally presented), the snapshot dataset was combined with the actions data to get a complete sense of the demographics of any individuals employed by the County for any length of time during the fiscal year. If a data point is based on one single fiscal year (e.g., median tenure of County employees) the fiscal year is noted in the text of the report.
- Hiring Report – We tested each of the hiring types (regular hires – grouped, temporary hires, regular hires only, temporary to regular hires) to understand the likelihood of hire by department, disability status, division, employee level, gender identity, generation, sex (ternary groups – female, male, non-binary), sexual orientation, race and ethnicity, tenure, union, and veteran status. The actions dataset was used for these analyses. Additionally, if an individual experienced multiple employment actions of the same type within the same fiscal year (e.g., hired as a temporary worker two times during FY 2022), they were counted as only one instance. However, if they were hired in two different fiscal years, the action was included for each fiscal year. Data for the personal demographics (e.g., disability status or gender identity) was provided by employees in their Workday profile. Not all fields are completed by all employees, resulting in high levels of missing information for disability status, gender identity, sexual orientation, and veteran status. Comparisons to hiring pool and Multnomah county community demographics are based on the 2020 Census and American Community Survey updated estimates. These comparisons are shown on the quarterly employment trends hires dashboard.
- Separations Report – We tested each of the separation types (voluntary, retirement, no fault and involuntary) to understand the likelihood of separation by department, disability status, division, employee level, gender identity, generation, sex (ternary groups – female, male, non-binary), sexual orientation, race and ethnicity, tenure, union, and veteran status. The actions dataset was used for these analyses. Additionally, if an individual experienced multiple employment actions of the same type within the same fiscal year (e.g., two voluntary separations during FY 2024), they were counted as only one instance. However, if they had separation actions in two different fiscal years, the action was included for each fiscal year. Data for the personal demographics (e.g., disability status or gender identity) was provided by employees in their Workday profile. Not all fields are completed by all employees, resulting in high levels of missing information for disability status, gender identity, sexual orientation, and veteran status.
- Other Actions Report – For promotions, demotions, work out of class or temporary assignment, limited duration assignment, reclassification, and lead status we looked at all employees who had at least one instance of the action during FY 2022 – FY 2024 to understand the likelihood of the HR action by department, disability status, division, employee level, gender identity, generation, sex (ternary groups – female, male, non-binary), sexual orientation, race and ethnicity, tenure, union, and veteran status. The actions dataset was used for these analyses. The actions dataset was used for these analyses. Similar to hires and separations, not all fields are completed by all employees, resulting in high levels of missing information for disability status, gender identity, sexual orientation, and veteran status. Due to the lower frequency of these actions occurring, analyses (beyond descriptive information) were performed for the three fiscal years combined. If an individual experienced multiple employment actions of the same type (e.g., two voluntary separations during FY 2022 - FY 2024), they were counted as only one instance.
Notes about specific data fields and calculations:
- Retirement eligibility was calculated using age, tenure at the county, police/fire status, and PERS tier. We did not have information regarding years of service earned outside of the county. Therefore, our estimates of retirement eligibility are likely conservative. Retirement estimates are shown on the quarterly employment trends separation dashboard.
- Generation was calculated based on employee date of birth using Pew Research categories for generation.
- Tenure was calculated based on County length of service.
- Employee action groupings (e.g., what was included as a voluntary separation) and employee level groupings (e.g., which job profiles were included for frontline supervisors) are available on the spreadsheet: Workday Level and Action Groupings.