Databases are not my thing. I understand them, but this is beyond my scope of figuring it out. I reached out to DataRepublican on X to see if she would be interested, but I did not hear back.

A tremendous amount of information is available online concerning the 2,313,216 federal employees on the payroll as of Sept. 2024. (There were 2,181,106 in Sept. 2020.) The FedScope data is posted online here - https://www.opm.gov/data/datasets/

The information is broken down into agency, location, age, education level, grade, length of service, occupation, occupation category, pay grade, salary level, supervisory status, work schedule, work status, salary and more.

In the dataset, you have the main file (about 165MB text file) and 17 supporting text files. The data seems to be updated about three times per year and we have information going back to at least 2000, but the data collected may vary.

The information in these files do not include the USPS employees or active-duty military. (As far as I can tell.)

Someone (elsewhere) had mentioned that the US Fish and Wildlife had "laid off" two employees in Florida who were (most likely) working in Wildlife Refuge Management. In their opinion, it was a disaster to lay off these employees. So I went digging. In Florida, the US F&W Agency's EE count went from 273 in March 2022 to 323 in Sept. 2024. Specific to Wildlife Refuge Management, the EE count went from 18 to 22 in those two years. (A 22% increase in staffing.)

I was also able to (manually) determine that 14 of the 22 employees in 2024 were "supervisors." I was also able to determine average pay and individual pay right down to the specific employee. (No personal info is available in the databases of course.) I also was able to (manually) determine the total payroll for departments. For example, the total payroll for the Wildlife Refuge Management area in Florida went from $2 million (18 EEs) to $2.63 million (22 EEs) in two years.

I'm not saying the data I found is a good or bad thing. I just think we should have an easy way to compare this information over time to measure the increase or decrease in the size of the US government, all the way down to the specific agencies and roles. Although I noted the increase in the size of the Interior Department's US F&W Agency (Headcount went up 7.5%), I did note the Department of Education's headcount has remained steady in the past four years.

What tools would be a good choice to make it easy for website visitors to crunch this data? For example, comparing EE count over time in a specific state, agency and/or department.

Is this even a worthwhile endeavor?

It's not like I have any free time to do this, but I already spent five hours or so going through the information manually, so I'm kind of invested in at least asking if a solution exists. All the data is readily available, but there does not seem to be a good tool to crunch it and display it.

Import the text file into Excel.

Use Excel pivot tables and other data analysis tools.

quote:

Originally posted by 4MUL8R:
Import the text file into Excel.

Use Excel pivot tables and other data analysis tools.

Thank you. With 2.3 million lines in the TXT file, unable to import the file (too much data). I went in and copy/pasted specific rows in a particular agency to do what I did so far.

I wonder if you could email DODGE asking to provide this somewhere to show what they are doing. Probably won’t go anywhere, but worth a try. I would think they would want to post that kind of data.

quote:

Originally posted by StorminNormin:
I wonder if you could email DODGE asking to provide this somewhere to show what they are doing. Probably won’t go anywhere, but worth a try. I would think they would want to post that kind of data.

They are sort of posting data from a year ago here - https://doge.gov/workforce

But they have no tools yet to compare over time or drill down to specific roles. I've asked them to post more details.

I don't know how to do it but I actually think it's simple to do with a recursive web database tool.

I'm not a coder but I took a masters program for information technology. In one class, the end of course project was to create an online database reporting tool.

I did my project by patching together codes from my other classmates; I was clear on what I wanted to do: just create a forecast from some actuals and you just select the part.

The weird thing is that only my program work and this other guy. Everyone else's in that class of 12 had issues during their presentation.

But my point is the other guy's program. He was working in banking at the time and his program impressed even the instructor. The program allowed him to analyze anything and I mean anything off the database. He should the program and it was just a few lines of codes.

The idea of the code is you enter what you want to analyze or pull and the code "generated(?)" the code to pull the data you wanted to look for and analyze.

Not understanding the code but seeing how few the lines were, I certainly knew it was elegant. And this was before AI; around 2005.

2.3 million records will require use of "R" software, or JMP, or SAS. JMP and SAS are very expensive annual licenses (JMP is $1400).

quote:

Originally posted by 4MUL8R:
2.3 million records will require use of "R" software, or JMP, or SAS. JMP and SAS are very expensive annual licenses (JMP is $1400).

That is what I've learned.

This is not a project I can afford to put together. Hence, I'm hoping DOGE or DataRepulican will run with it. It would provide so much insight.

Thank you.