State Payroll Database
Inspired by my experience as a former state employee, I am interested in research questions related to the extent to which states offer compensation that is competitive with the private sector and how successful states are in retaining highly-skilled employees.
The map to the left shows the states shaded in blue from which I collected state employee payroll data.
While some states post payroll data on budget transparency websites, many require Freedom of Information Act requests made to the data custodians in order to receive the data.
For most states, the dataset includes a ten-year panel of payroll for individual employees of each agency, totaling about 36 million panel observations, which includes about 94% of all state employees in the United States.
Click here for a spreadsheet that shows the availability of years and variables by state. The spreadsheet will be updated as I receive new data from states. I am currently expanding the dataset to include the latest year available (generally, 2023 or 2024).
If you have a research idea for which the payroll data might be useful, I invite you to contact me by clicking the "Contact" link above. I am open to embarking on new research collaborations.
Example Uses in Research
In "Cost of Living Adjustments and Public Workforce Turnover," we are interested in whether annual cost of living adjustment (COLA) freezes as a budget balancing tactic explains turnover among affected state employees. We use annual changes in the total pay variable for full-time civil service employees to observe cases where a COLA freeze appears to occur. We then observe the turnover rate of employees in affected states and make comparisons to states that did not pursue a COLA freeze. This is a case where every agency in every state is used to calculate the rate of COLA increases and turnover rates.
The panel structure of the dataset is used to calculate turnover and retention rates used in "Organizational Memory and Snap Back Performance in Public Agencies." In states whose panels begin on or before 2012, I identify individuals employed in state health departments in 2020 with at least eight years of experience. The goal is to measure whether agencies with high retention rates of long-term employees were better able to develop strategies to vaccinate their adult populations for COVID-19, presuming long-term employees possess organizational memory from previous crises.