Politics of Professions: More Fun With Partisan Graphs

Maybe you’ve seen this very cool graph showing the “politics of professions”?

Here’s a screen grab of a small part of it:

Politics by Profession 1

Below these amusing comparisons is a set of little colored circles for each considered profession that you can click on and expand:

Politics by Profession 2

These are really fun, and largely confirm a number of biases. Here, we see that the Media and Academia are leftist strongholds, or that most people who can think shun the right – take your pick.

But, alas! What you think you’re seeing – what the title “Which way does your occupation lean?” suggests you are seeing – you are not seeing. Once again, just a tiny amount of curiosity and research reveals that this lovely, award-winning graph doesn’t say what it claims to say. And it’s the usual suspects that render this graph, like so many before it, largely meaningless:

Method. Well? How did the authors arrive at these numbers? Nowhere stated – which, to put it mildly, is not a good sign.

Data source. Where did these numbers come from? As highlighted in the title to the first screengrab above, “campaign contribution data from the FEC”. Googling that directs one to the FEC website. There’s a lot of stuff on that site, I don’t have all day to sift through it, so here are some preliminary considerations:

  1. Are we only considering those who have made political contributions in some manner to one of the two major parties? And in a manner that requires the FEC keep track of it? This means if I’m a mathematician, for example, who votes consistently for one party or the other but can’t bring myself to part with any cash to fund them, I don’t count? Isn’t that *most* people, or at least a huge percentage of people?
  2. What period is covered? The database runs from 1997 to approximately the present. The same person – the Donald, for example – might change who and what he supports over time. Sticking with this example, he might be a staunch and generous Hillary supporter right up until he decides he needs to be a Republican President. Is this accounted for? How?
  3. An awful lot of people in business will give to *both* parties.(a) If I’m an investment banker, I might largely vote Republican but *always* contribute to Democratic causes – it’s just simple prudence. It’s just business. Investment Bankers, according to this graphic, are split right down the middle – I strongly suspect that this is not the result of a neat and fundamental split in partisan loyalties right down the middle as it is a result of a sort of natural selection: investment bankers would be idiots NOT to make sure that candidates from BOTH parties get a cut. Is this addressed, somehow, in the graphic?
  4. If I give $100 to Democratic causes and $1,000,000 to Republican causes, is that accounted for?

Classifying Professions. Who determines the list of professions? Who decides where the line falls between, say, a mathematician and a statistician and a data analyst?

  1. Many people have multiple professions, or their work straddles a number of professional classifications. (Am I a strategic planner? A sales manager? A market researcher? A content creator? A pricing analyst? A product manager? In my current job, the answer is ‘Yes’.)  Is that addressed? How?
  2. Are these claims verified? In other words, does the person donating to a partisan cause get to just say what their profession is? If I say I’m a rocket scientist, do I get any push back?

This is similar to the problem of asking people in exit polls to say what their highest level of education is – the guy with the degree from a barber college can say he is a college grad, while the gal who dropped out in the 6th year of a nuclear physics degree might say she isn’t. Even if they answer truthfully (and what would make them?) the answers are not giving you a true picture: that college grads (assumed against most of the evidence to be the smart people) voted for X, while those with no college degree (assumed to be the dumb people – you know, like Gates, Jobs, Wozniak)  voted against it. So if I ask you your profession, and you once had a bit part in Galaxy Quest but now wait tables and live with your mother, you put ‘actor’ as your profession, right? Or if I got a government grant once to try to sell small towns on putting in solar panels, I’m an entrepreneur, right? Me, I’ve helped people in far-away places figure out the math involved in leasing an airplane – so I’m an international financier, right?

It’s fun to imagine that this set of little graphs tells us, for example, that 81% of roofers are Republicans while 80% of philanthropists are Democrats. But that is not what it is telling us. What it is telling us, most likely, is that 80% of the donations by count (not amount) made by people who a) made political donations tracked by the FEC since 1997,  and b) self-reported that they were philanthropists went to candidates/PACs/causes/whatever that the FEC classified as Democratic.

Maybe. The lack of any explanation other than the general statement that the data came from the FEC inclines me to assume the most simple approach was used: add ’em up by claimed profession across all the years, divide the Democratic total by the gross total to get a percent, slap a graph on the numbers, and voila! You have a hot steaming pile posing as information. Could be wrong, but don’t know how I’d find out.


a. If you think that one of the major purposes of all this FEC disclosure isn’t so that the elected officials can check up on to whom they owe favors and to whom smackdowns, you need to read up on how, say, LBJ operated. He *always* kept score, and knew *exactly* what he owed and was owed from everybody he worked with, down to cub reporters. The essential part of implementing Callicles’ model of political virtue (Punish your enemies, reward your friends, indulge your every whim) is knowing who is a political enemy and who a political friend. That, and, I suppose, cultivating memorable whims.

Author: Joseph Moore

Enough with the smarty-pants Dante quote. Just some opinionated blogger dude.

One thought on “Politics of Professions: More Fun With Partisan Graphs”

  1. Well that’s always the problem with studying humans, it’s the same as the coastline paradox (which is why I say, even if you’re not religious a person should be able to grasp that an individual is still larger than any government could ever process).

    But if you go look at OpenSecrets a lot of its data seems to confirm this one.

    And I did find the original’s statement on method. They freely admit there would be obvious issues with it, though I think they’ve done good work and until a better metric is devised it’ll have to do.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s