Welcome! If you’re reading this, you’re probably one of the data-loving folks who clicked over from my article on income data for young journalists. I wanted to put together a separate piece on my methodology, because I was interested in sharing exactly how I put together the data.
Just the Facts, Jack
You can get anonymized CPS microdata for your own purposes from the University of Minnesota. My first step was to put together an extract for March 2011-2014 CPS samples (note: this is 2010-2013 data, which I selected specifically to avoid the recession). I added some important variables: Wage and Business Income data, which captures writers working for themselves (say, like bloggers and other writers for small sites), educational attainment, usual hours worked and age.
Next, I had to update all of the income data for inflation – all dollars have been normalized to the 2013 data using the methodology in this link. I also added together wage and business income – this would capture folks who both run a web site or publisher and make money, as well as collect a paycheck. That’s simple – just make a new column and add your inflation adjusted income data together.
I then narrowed my data down by only counting people who usually worked at least 1 hour weekly in that year. I also filtered the population to only contain folks who had attained at least a Bachelor’s (EDUC >= 111).
The Subjective Decisions – Direct Your Ire Here
(Actually, We Prefer You Ask Nicely If These Aren’t Clear)
What is ‘Young’?
I ultimately settled on 34 years old because of a throwaway line by Felix Salmon – he mentioned he had been in the industry about 13 years before his break. 34 years old is about 11 or 12 years after a four year degree, and high single digits for more years of college. Additionally, it’s about exactly 1/3 of the way through a 40 year career – so if you like breakpoints, maybe that’s where ‘junior role’ becomes ‘role’ on its way to ‘senior role’. (At least mathematically, since ‘junior’ probably drops from a title well before that!)
Anyway, the youngest age in the data set was 20 years old – so there are still some very high achievers in there.
How to Decide What Roles Journalists Might Hold
I really disliked the “determining what to count as a journalist” part, because I’m not in the industry and I don’t want my subjective judgement to color the findings. That said, I came up with a methodology to match writers, but also to screen by industry – a decision I felt that would best represent editors and writers working in a journalistic capability. Here’s those screens, which I feel are a fair representation of journalists:
- OCC: (2810) News analysts, reporters, and correspondents (2825) Public relations specialists (2830) Editors (2840) Technical writers (2850) Writers and authors
- IND: (6470) Newspaper publishers (6480) Periodical, book, and directory publishers (6670) Radio and television broadcasting and cable subscription programming (6672) Internet publishing and broadcasting and web search portals (6680) Wired telecommunications carriers (6690) Other telecommunication services (6770) Libraries and archives (6780) Other information services
Obviously opinions differ here, but I can tell you I really sweated four additional industries in particular: (9160) Religious organizations (9170) Civic, social, advocacy organizations, and grantmaking and giving services (9180) Labor unions and especially (9190) Business, professional, political, and similar organizations
I ultimately decided against including those four choices, but I can be persuaded to rerun the journalism numbers again with them included (or any other industries there is a good argument for, actually).
After OCC I had about mid-1600s of data points, and dropped to 400 and change after the industries, so it’s still a reasonably fair sample.
I’m not convinced adding or dropping any industries will make a significant difference or bridge (or increase) the gap, but I’d appreciate your comments either way.
The Data and Sample Sizes
The only ugly sample size was for 18-34 year old journalists with a four year degree; we had a mere 13 data points per quantile. You can almost definitely count the median (remember, the CPS is a weighted data set) and the numbers surrounding it as pretty good, but don’t take the exact numbers as the perfect truth. Here’s the full count:
|Income Quantile (Wage and Bus. Inc.)||8.33%||16.67%||25.00%||33.33%||41.67%||50.00%||58.33%||66.67%||75.00%||83.33%||91.67%||Samples|
|All Journalists With Bachelor’s||$11,172.41||$25,916.47||$32,000.00||$35,548.59||$41,466.36||$48,722.97||$57,016.24||$67,382.84||$74,818.52||$85,000.00||$115,000.00||421|
|All Working Adults With Bachelor’s||$10,366.59||$21,200.00||$30,996.24||$38,905.63||$45,959.95||$53,441.80||$62,199.54||$72,680.85||$85,506.88||$105,000.00||$148,568.20||370493|
|Journalists <= 34 with Bachelor’s||$6,000.00||$21,376.72||$26,953.13||$31,099.77||$33,500.00||$36,283.07||$40,429.70||$42,503.02||$47,000.00||$57,016.25||$75,000.00||142|
|All 18+ Workers <= 34 With Bachelor’s||$9,329.93||$17,625.28||$25,000.00||$31,000.00||$36,340.42||$41,466.36||$47,686.31||$53,441.80||$62,199.54||$74,818.52||$96,195.24||121539|
I think that covers everything, but if I missed an explanation please let me know.