It has been a while since I last began playing around with data.world, an awesome collaborative website where users can share data sets to draw insights.
I chose to take public compensation records from the City of San Francisco, mostly because public employee data is one of the easier bits of information to get a hold of.
To be honest, I really have no hypothesis or operating assumptions – I’m just seeing what happens, so to speak.
The first thing I did was download 2013-2017 San Francisco employee compensation records from data.gov, a US run website created in 2009 as an effort towards transparency. Fortunately the information was pretty clean, so no scrubbing was necessary. I then imported the CSV into data.world site as a new project. Data.world has a very cool query builder that allows you to leverage SQL.
I noticed that the most standardized field for categorizing job group was rolled up to a ‘Organizational_Group’ , which fit the entire data set in 6 buckets
|Public Works, Transportation & Commerce|
|Culture & Recreation|
|Human Welfare & Neighborhood Development|
|General Administration & Finance|
I thought it would be useful to see the average salary of each organizational group, here is for 2017
|Year||Organization Group||Average Salary|
|2017||Public Works, Transportation & Commerce||$104,978.49|
|2017||Culture & Recreation||$50,408.47|
|2017||Human Welfare & Neighborhood Development||$71,870.05|
|2017||General Administration & Finance||$97,371.90|
I thought this was kind-of useful, but it would be cool if I could see what the average salary was for each organizational group, year-over-year. I published the following ‘insight’, which can be shared with followers on data.world
It looks like while public protection enjoyed, on average, the higher salary or ‘total compensation’, which includes benefits – it had slightly dropped year over year, compared to Community Health, Public Works, and General Administration & Finance which has been growing slowly over time.
My next question was based on the total amount spent on employee compensation by organizational group, year after year.
Here is what I found:
Okay, okay… that was done in excel…
THIS, however was another insight published on data.world
This insight, I hope, shows that each organizational group has taken about the same percentage of the pie year over year, with the most drastic decrease for Public Protection and an increase in Community Health.
What have I proven? Probably nothing, but it was fun to play ‘data scientist’ for a bit!
Make your own profile and follow me on data.world!