Conservative vs. liberal jobs

Robin Hanson reports:

My last post got me thinking about the liberal vs. conservative slant of different jobs. Here are two sources of data.

Consider some jobs that lean conservative: police, doctor, religious worker, insurance broker. These seem to be jobs where there are rare big bad things that can go wrong, and you want to trust workers to keep those from happening. That explanation can also makes some sense of these other conservative jobs: graders & sorters, electrical contractors, car dealers, truckers, coal miners, construction workers, gas service station workers, non-professor scientist. Conservatives are more focused on fear of bad things, and protecting against them.

Now consider a set of jobs that lean liberal: professor, journalist, artist, musician, author. From these you might focus on the fact that these jobs have rare but big upsides. So the focus here might be on the small chance that a worker will be come a rare huge success. This plausibly seems the opposite of a conservative focus on rare big losses.

But consider these other liberal jobs: psychiatrist, lawyer, teacher. Here the focus might be just on people who talk well. And that can also make sense of many of the previous list of liberal jobs. It might also makes sense of another big liberal job: civil servant.

I’m not suggesting these are the only factors that influence which jobs are liberal vs. conservative, but they do seem worth exploring.

Which other factors might help explain the distribution of conservative vs. liberal jobs?


Performance vs Presentation?

"psychiatrist, lawyer, teacher": barriers to entry applied by the state.

Once upon a time psychiatrists also listened well.

dearieme, almost any occupation has professional qualifications which you can characterize as a barrier to entry created by the state. But this doesn't mean they are liberal or conservative for that reason. Doctors, for example, or nurses, or investment bankers who have to take SEC exams required by their employers, engineers who need certifications, or an insurance agent or broker who has to take exams or qualify to be licensed.

I don't buy it.

Yeah, but with engineers, doctors and nurses those barriers are more justified, the purpose is to prevent a bridge from collapsing or a patient from dying.

That is certainly one purpose. Another purpose is to limit competition. This is a classic case of bootleggers-and-Baptists.

Clover, The analysis attributing political contributions to occupations is based on employer contributions to PACs and director contributions to political parties. So, if you are a tobacco company which contributes to conservatives, your employees, ergo, are conservative.

The dichotomy you are constructing is just as strained I think. Auto dealers are listed as conservative based on their political contributions. Is that because they are conservative, or because they are contributing to conservatives to block specific auto dealer lending regulations. Same with the Exxon/Mobil workers which are, using the data based on the employer PAC contributions and director contributions, are they more conservative, or are they just contributing to conservatives who are blocking industry specific regulations.

I think you know the answer.

Clover, you don't think it's disruptive to have unqualified lawyers giving people terrible legal advice? The law is one place where strict ethical standards seem to pay off.

Either way you haven't really shown a correlation between necessary or unnecessary licensing and political leanings.

@ Aaron Luchko

If passing a bar exam makes one capable of dispensing valid legal advice why are there such disparities among attorneys? Why is Eric Holder in a better spot than the public defender in Sedalia, Missouri when both have been certified?


For the same reason that not every doctor or engineer is equally capable.

It's all but impossible to objectively measure the quality of a psychiatrist's services, and not much easier to objectively measure the quality of a lawyer or teacher.

This is in contrast to (for example) a design engineer, where the thing designed either works well and reliably or it doesn't.

Could it be that part of the divide is between jobs where the quality of the work product can be objectively measured, and ones where it cannot?

It's as easy to objectively measure the quality of a psychiatrist's services as it is to measure an interist's or dermatologist's or really any medical specialty where the focus isn't on procedures with largely binary outcomes. And looking at how well most products are designed, it appears to me that the feedback loop in that field isn't sufficient to weed out mediocre design engineers. Working well/reliability are almost never binary, but are on a spectrum. Anyone who has ever bought a car or a computer or an appliance, or really just about anything, knows this.

As for dearime's suggestion that liberals populate fields with state-applied barriers, I'm not sure how " professor, journalist, artist, musician, author" fit into that cubby hole.

It’s as easy to objectively measure the quality of a psychiatrist’s services as it is to measure an interist’s or dermatologist’s -

I have residual acne on my face or I do not. My displastic nevus was excised without pain or infection or it was not. Crap like 'depression' is evaluated through instruments designed by mental health tradesmen themselves.

That's not true of doctor, electrical contractors, car dealers or insurance broker?

It seems to me that a big differentiator here is that the most of the liberal jobs are primarily employed by the government, while most of the conservative jobs are employed by the private sector.

Another potential factor may be that many of the liberal jobs require high education but receive relatively low pay.

"Another potential factor may be that many of the liberal jobs require high education but receive relatively low pay."

Right. Resentment plays a big role in ideology.

I think the people that embark on SocialScience degrees aren't expecting to ever get rich.

But they feel they *deserve* to get rich. Or at least the people who do get rich don't deserve it.

Image enrolling at an Ivy level school with a 4.3 HS gpa and 1450 SATs, getting a Political Science AB with a 3.9 gpa, and then spending 1 year at Gawker as an unpaid intern, followed by 1.5 years at Conde Nast as an unpaid intern, then getting a full-time job at Buzzfeed for $3000/month. Oh, and you have to live in NY/DC/LA/London.

Meanwhile, that cisgendered, heteronormative, rethuglican non-hispanic pale male (who got in with only a 3.5 gpa b/c he's a Lax Bro) that graduated with a 3.2 in Econ has been making 6 figures at MEGA BANK since graduation (a Frat Bro hooked him).

actually, gawker pays its interns.
very doubtful that those writers believe they deserve to get rich. but some question the arbitrariness of success for some of the professions where the value add is questionable. there is a distinction.
also, lol @ "having to" live in NY - most people wouldn't have it any other way, whatever their standard of life appears to be to someone in the burbs with 2500 sq ft and a >30 BMI.

But they feel they *deserve* to get rich.

What do conservatives feel? How come conservative media, both broadcast and online is littered with ads for various get rich quick schemes and supposedly secret ways to milk the gov't for higher social security checks?

Boonton: Conservatives *want* to get rich. Liberals *deserve* to be rich. Ergo, conservatives suck it up to try and meet the demands of the market. Liberals fight the market and seek government override power.

That's funny as many ads I see on liberal media are for 401K services, mutual funds etc. The 'patent your invention' type ads seem equally dispersed. But 'get extra social security checks' seems to be squarely on conservative media.

How is that 'meeting the market'? How is that motivation the supposedly more innocent 'want to get rich' as opposed to the evil feeling that they 'deserve to be rich'? If a liberal feels he deserves to be rich because he spent a lot of time investing in his skills how is that different from a business owner who feels he deserves to be rich because he spent years building up his business?

"Liberals fight the market and seek government override power."

I guess that's why all those musicians, writers, and artists are organized into powerful blocks to seek government overrides to make themselves rich.

@Boonton: The ads that you see on various websites are selected because of your web history, in almost all cases. Websites can choose categories, but usually not specific ads.

"I guess that’s why all those musicians, writers, and artists are organized into powerful blocks to seek government overrides to make themselves rich."

You mean they join the Democrat party as would-be wealth confiscators? Well, they do what they can I suppose.

@Boonton: The ads that you see on various websites are selected because of your web history, in almost all cases. Websites can choose categories, but usually not specific ads.

Actually I just deleted yet another National Review email "Trick adds $155,525 to your Social Security..." This is not from my browsing history (if it was then it would be just as likely to show up on non-conservative sites and places). It has been well documented that a lot of conservative outlet ad revenue is devoted to scams, schemes, and frauds. So why is it not fair to ask about the character of the outlet's main audience?

@Boonton: "Trick adds $155,525 to your Social Security…" is about how to utilize the options made available to most social security recipients in order to maximize total revenue from the program. It is not a scam, scheme, or fraud. Usually this information is a segue to selling financial planning services

Oh, Scott, it must be tough, keeping one step ahead of all those parasitic artists, writers, and musicians who are trying to extract your precious bodily fluids! I imagine you as Rick Grimes, keeping your Colt Python loaded and ready for those wealth confiscating zombies who pop up every time you let down your guard.

Boonton: Sounds like a neat trick, I mean what's so hard about printing money right?

Scott, it does certainly seem to be structured to appeal to someone who is just as happy to scam (provided he won't be caught and punished) as to simply 'take advantage of all the options offered'. Yes yes I'm sure at the end of it all is something like 'financial planning services'. After all the company has to provide something to those it suckers into actually buying. But how exactly are they trying to hook their customers in conservative media? By appealing to selfish emotions, especially selfishness that comes at the expense of other people.


Interesting point. Most liberals who argue that right now the monetary policy of the US should be more loose do not seem to generally think they will get the 'money they want to print'. That accusation was made here that liberals take jobs where they may not earn a lot of money but think they deserve a lot of money...the presumption is that they support income redistribution in order to get their hands on that money!

But perhaps this is the pot calling the kettle black. Most liberals who advocate things like extended unemployment, enhanced foodstamps, Medicaid, even stimulus infrastructure projects are not generally unemployed, uncovered, hungry or even work in a construction field that would most directly benefit from stimulus. Perhaps the difference here is that conservatives are more comfortable with the position of 'I'm going to get everything I can for myself'....hence demand lower taxes and order the book on how to scam Social Security into doubling your checks! The accusation that liberals just want to redistribute money to benefit themselves, then, comes from their inability to understand a desire to help the general public...even if they may mistaken about the policies that would be the most benefit.

@ Kevin: Please make up your mind whether you want to deny the reality or defend the reality. There is a difference.

@Boonton: It's really sad. Successful liberals are just dying to give all their money to the poor, but the bad conservatives won't let them. If only we lived in a free society.

If they do decide to try to get rich, they go to business school. (I did; I'm not rich, but I'm doing fine.) Then they become conservative.

Most B-school matriculants have a social science background, more often than not economics.

My far left "friends" and acquaintances are deluded. They are cranks and radicals, who wish to shape the world so as to make it line up with their utopian hopes and aspirations. They are not resentful.

". . . lean liberal: professor, journalist, artist, musician, author. . . . " Right. They involve doing stuff that can be fabricated.

Mike, the data is based on political contributions by corporations and PACs. This is not data based on the contributions of individual employees. See my link to the source of the data below.

Here's a graph of individual donors by industry:

Steve, This data is based on PAC contributions.

Bill, they categorized them by industry, but studied individuals.

tmc, no. that's not true. read the links and the data description in table listed in the links below.

Before the "Citizens United" decision, corporate PACs were funded primarily through employee contributions.

The corporation still decided where the funds were going, so it isn't a perfect proxy, but who would put money into a corporate fund that habitually goes to their political opponents?

Nobody, IF putting money into the corporate PAC was really "voluntary". I never felt that it was really "voluntary" as one would usually understand that term. A few bucks sucking up can be a good investment.

I've been asked to donate dozens of times, no one with any sense thinks it's voluntary.

I can recall having it suggested to me to kick back a small percentage of my salary into my corporate employer's PAC. It was done pretty apologetically. I can't recall if I did it during the short time I worked there.

When my wife worked for Cook County as a COBOL programmer, she found out she was pretty much required to kick back a more sizable percentage of her salary into her boss's boss's boss's re-election campaign fund. The pressure wasn't subtle.

So, it's hard to say how voluntary these kind of contributions are.

When Alphonse d'Amato was the principal enforcer of the Margiotta machine's shake-downs, county employees were expected to kick back 1% of their wages and salary. That included the faculty at the Nassau Community College.

A more encompassing way to say this might be that conservative jobs are more likely to have performance that can be measured in some concrete terms, while liberal jobs are typically harder to measure that way. (This has some overlap with the "talk versus do" division.)

For example, a building contractor makes a building that does or does not satisfy the contract, does or does not meet building codes, etc. And of course anyone in the private sector has a collective measurement by whether the company is making a profit.

Even a policeman's job, though more difficult to measure, can be looked at through metrics such as arrests, complaints, and collectively through crime rates, etc. And similarly farmers by their yield rate, etc.

There is much more tenuous link between, say, a professor's performance and course enrollment, or the university's budget. And public school teachers resist most adamantly any link between how their students do and any judgement on their performance. To say nothing of trying to measure the performance of a civil servant, where the supposed goal of the employing agency (EPA protecting the environment, for example) may not be the same as the bureaucratic goal (defining more and more things as needing protection).

A more encompassing way to say this might be that conservative jobs are more likely to have performance that can be measured in some concrete terms, while liberal jobs are typically harder to measure that way. (This has some overlap with the “talk versus do” division.)

Hollywood actor seems pretty concrete as there is no shortage of objective measures of what movies or tv shows make a lot of money and which don't. On the flip side doctors are surprisingly difficult to measure objectively. Many doctors are 'small businessmen' running their own practices but it is very hard to measure how well they are doing as busienssmen against other docotrs. Yes if they do very badly they will go out of business but my experience is many doctors offices are run very poorly as busiensses but since the rent gets paid and payroll is run every other week there is nothing stopping them from being pushed out of the market.

"Hollywood actor seems pretty concrete as there is no shortage of objective measures of what movies or tv shows make a lot of money and which don’t."

It's much harder to figure out exactly the contribution of one specific actor to the profit margin however; even in the case of one single Tier 1 name in the movie. (Of course, given the accounting practices of Hollywood, it's hard enough to actually know what the real profit is anyway.)

It isn't hard at all. There's a reason why people like Pitt or Clooney command top dollar. When it comes to music it is even more clear cut. Taylor Swift's downloads and CD sales can be easily tabulated. Yet this is a group that veers overwhelmingly liberal which pretty much defeats the idea that liberals are people who have jobs that can't easily be evaluated by objective criteria. In terms of market structure, the artistic markets are about as close to pure Ayn Randism as you can get (which is probably why she was so attracted to Hollywood).

"Most conservative jobs employed by private sector" - like "police" the first job named? Military is also known for being pretty happy to vote GOP.

These comments mainly show that people know what motivates people who vote for their party, but are clueless when it comes to the motivations of the other side.

That said, I will try and guess the motivations of GOP leaning professions - many seem to involve working in very hierarchical places with clear lines of authority (military, large corporations, churches).

The other difference seems to be tolerance of ambiguity. Those seeking black and white goals (maximize sales) go GOP. Those that enjoy grays (what is a good song?) go Democrat

Never underestimate the power of random variation and path dependence. Suppose some of these jobs are more liberal/conservative by coincidence. However, these differences persist as people self-select into jobs where people share their values and beliefs.

+1 Path dependance is too often neglected (he says via his QWERTY keyboard). Path Dependance and US healthcare financing anyone?

I don't think it is really an accurate characterization to say that a religious worker, the most conservative profession, is primarily someone who prevents rare big bad things. That sounds more like an outsider's caricature. Some other jobs just don't make any sense. Are graders, electricians, and truckers motivated by fear of bad things? Non-academic scientists and academic ones do similar work, but in different contexts. One is conservative and the other is liberal.

This "conservatives are motivated by fear" thing is irksome and not really supported. It is the kind of argument made by those who don't take conservative ideas seriously and thus feel the need to explain conservatism psychologically (not saying that Hanson is actually in that camp).

Also, many of the careers simply have self-interested reasons to support one party or another. Conservative policies tend to support law enforcement more, liberals want to spend more on education, liberals tend to support more restrictions on construction, etc.

Agreed. It also depends on the affiliation of the religious workers. For example most mainstream protestant religious workers appear empathetic and support liberal government policies.

dan111, The data on which this is based is PAC contributions by the employer, not the contributions of the employee (unless a union).

PACs collect donations from individual employees. Corporations themselves cannot donate.

yea, religious worker is by definition someone who "talks well."
obviously a biased characterization by hanson.

I can see religious counseling is often predicated on preventing big, bad, but rare things. Most religions encourage people to live more modest lifestyles, take more precautions, save more etc. That does have the effect of making people more prepared to handle 'big bad things', even though they may never come to pass. That doesn't mean the actual theology of particular religions is premised on being some type of self-help philosophy for living a better material life.

This “conservatives are motivated by fear” thing is irksome and not really supported

Indeed, consider the Civil Servent. Isn't that job often motivated by fear of a 'big bad thing'? 99.99% of the time no one will notice what you do all day long but if a building collapses that you signed the permit on, if someone you let out on parole kills someone, if some policy erupts into an Internet firestorm, you could be singled out! So play it extra cautious no matter what.

"if a building collapses that you signed the permit on, if someone you let out on parole kills someone, if some policy erupts into an Internet firestorm......"

No repercussions. Gov't immunity and all.

Preventing "big bad rare things" is a concern of many religions and religious people; the problem is seeing that as the primary concern and the way religious workers see their own roles.

Not really immunity. Finding someone to put all the blame on happens a lot in gov't work. In contrast if you were a top energy trader for Enron you probably have not spent the last few decades unemployed. If you were head quant for designing MBS's, I doubt you are living on extended unemployment today.

In contrast in gov't employment you can be scapegoated even for doing your job correctly. Back during Rudy's term as mayor of NYC there was a guy who spent about a week in jail on some minor charge, the prosecution couldn't get the evidence together to make a case, though, so the judge ordered the guy released. Totally by the book. Guy went out and killed someone and the judge was pillored by the mayor, politicians and the press for letting the guy out to kill. Perhaps the judge didn't loose his job but his career prospects probably ended that day.

BTW, while we are talking about 'liberal' versus 'conserative' jobs, why is gov't employment considered liberal? Hello, active duty military and national guard? Leave the movies aside, most people in the military are doing what are essentially jobs and are not on SEAL Teams hunting down Bin Laden. I suspect career military people lean Republican but the nature of their jobs are probably very much like regular civil serve employment.

Also secondary point, Car Salesmen and Real Estate brokers. Highly Republican group but sorry no Teacher's union in ths history of the US has ever gotten the special gov't breaks these two industries have and yet most people aren't even aware of it.

Random variance. Ask yourself this: is there anything inherently more liberal about being a lawyer than an insurance broker? Is it three years of law school? Come on be serious.

Insurance broker is a sales job. Marginal tax rates matter in sales. A lockstep compensation system is virtually unthinkable in insurance brokerage. Not the case among lawyers.

Sales guys (and their wives) are a sizable chunk of Republican voters.

That would seem to argue against the thesis here that Conservatives are more pessimistic, fearing the rare bad event and liberals are more optimistic...hoping for the rare beneficial event. Selling insurance, of course, is premised on getting people to fear something rare but bad will happen. But many other sales guys have pitches premised on optimism (time shares, cars, real estate, etc etc.)

Right, sales guys have to be optimistic. There's a big industry of motivational speakers to keep them fired up.

Hmmm, so perhaps insurance is in the same boat.....'hey if your wife dies you'll be rich...or hey if you get disabled you'll get $5,000 pe rmonth!"....the rare bad event is flipped into an optimistic one. But that still goes against the thesis here that conservatives tend to be people who are trying to avoid rare bad events more than liberals given that salesmen are generally a Republican group.

Perhaps we should modify the thesis to say Conservatives believe both in preparing to deal with or prevent the rare bad event AND also have a gut feeling if you 'do the right thing' life will generally work out for you. Therefore people who are less fortunate are either blameworthy because they didn't prepare properly (buy insurance, save money in case of a layoff, etc.) and if they only start doing the right thing they will be fine. Likewise people who are doing well both did a good job preparing to avoid bad things AND were consistent enough doing the right things that they propsered. A play on the old Protestant Ethic.

In contrast Liberals are more open to the idea that misfortune can be random, like the Book of Job teaches, hardship can fall on the worthy just as easily as the unworthy. They are also more open to the idea of collective action to ward off bad things. For example, storing the surplus grain for 7 years to help get through 7 years of famine in the future. Conservatives IMO would be more comfortable with hearing about how you should hoard 7 years of food in an underground bunker and stock up on ammo to ward off your starving neighbors.

Grrr. You are supposed to hedge against the possibility that bad things can happen to you. Buying insurance is the easiest and cheapest way to do that. Insurance salesmen are preying on fear only if the insurance is not fairly priced. Competition usually makes sure that doesn't happen.

Keep in mind too....

The teachers' unions have nothing in terms of gov't favoritism when compared to car salesmen and real estate brokers. Despite an obsessive amount of feeding on the myth of 'self made men/women' 95% of these people would either be out of a job or be back office support if it were not for a slew of gov't policies designed to favor them. Ask yourself why are you unable to just go online and buy a new car from Wal-Mart or Amazon or directly from the manufacturer? Next time you buy a house look carefully at those closing costs and ask yourself what exactly did those people do for that money? What real risks are they taking for things like 'title searches/insurance' or listing a property?

The teachers’ unions have nothing in terms of gov’t favoritism when compared to car salesmen and real estate brokers. Despite an obsessive amount of feeding on the myth of ‘self made men/women’ 95% of these people would either be out of a job or be back office support if it were not for a slew of gov’t policies designed to favor them.

Let go of everyone's leg blowhard.

Benny, go look at the source of the data: it is corporate PAC contributions. See the link to my comment.

Law is a self-regulated profession, consistently among the least impacted by government regulation. There's less incentive to think seriously through the comparative costs of government expansion (especially when it creates more business for lawyers).

actors - very much alike the aforementioned jobs that shift your focus on "rare but big upsides". Consider the usual, accomplished TV oder movie star: all that he is praised for is his skill or looks - something he'd achieved a very long time ago. The big distance in time between ability-breakthrough and commerical-breakthrough could leave you with a strong sense of unefficient markets.

You could tell this was going to be an insipid post from the headline. And it did not disappoint.

>Conservatives are more focused on fear of bad things

Uh huh. No bias here at all. And given that the health insurance industry teamed up with Obama's government to thoroughly screw the American taxpayers for the profit of both, nice job listing "insurance broker" as a conservative job. Jesus, man.

Right. These days, political liberals tend to be more cautious and day to day conservative, while political conservatives tends to be more favorable toward risk and development. People who work on commission, for example, tend to the right, while people with tenure or civil service jobs tend to the left.

For example, Democrat Rob Reiner just got the nice liberal people of Malibu to vote for an initiative that in effect bans the building of a Whole Foods in the giant empty lot in front of the Malibu City Hall. The residents of Malibu like Malibu just the way it is. If their servants have to drive a long way to shop for them at a Whole Foods, well that's just the way it is and always has been in Malibu.

The only bad thing is the mangling of the data. This data which infers the political views of employees is based on corporate PAC contributions.

So, Target employees, by this definition and the data, are conservatives.

What a hoot.

Sorry, but you are misusing the data, and you should discuss what the DATA is first in the post.

The DATA is political contributions, PAC contributions by various groups. Here is the link to the DATA:

Guess what: Defense contractors contribute to conservatives. Guess what: Exxon Mobil gives to conservatives. So, you get to say: defense workers are conservative, and oil workers are conservative. Guess what: public employee unions (teacher unions) give to democrats, so you get to say: teachers are liberal.

The chart from which Hanson makes these claims is based on the text from the website above. Here is what the author, not Hansen, has to say about the data (notice the qualifier at the end):

"The second reason is that, despite its status as a Fortune 500 company, Google is among a growing group of powerful corporations closely aligned with the left. Based on my estimates, Google’s employees are the fourth most liberal of any U.S. corporation, behind Genentech, Apple Inc. and Starbucks (See figure above or click here to view in table format). A table with the To put this in perspective, consider that during the 2008 election cycle, Google employees raised $20,800 for John McCain and $55,451 for Ron Paul; compared to $89,300 for Hillary Clinton and an astounding $803,436 for Barack Obama. In fact, in terms of political contributions, the employees at these firms more closely resemble faculty at liberal universities than traditional Fortune 500 corporations. In some ways, this makes sense. With the exception of Starbucks, each of the most liberal firms, in large part, deals in research and innovation, and each has a reputation of actively recruiting Ph.Ds.

This is the where I would usually caution against putting too much stock into what the ideology of employees reveal about a corporation’s decision-making, because it is the board of directors that ultimately decides issues of this magnitude. However, Google’s board of directors also appears to be decidedly liberal."

The link is here:

In context, this is a warning against reasoning from individual contributions to corporate policy. In other words, it is a warning against the precise mistake that you are making over and over again in these comments when you claim that the corporation made these donations. Ironically, this is what you chose as evidence that others are misrepresenting the data.

It clearly says that the data is on individual contributions, not contributions by firms.

It clearly says PAC, and it clearly says that the contributions are by directors of the companies.

It is in the link and the statement at the end of the data clearly, but evidently not clearly enough for you, the following:

This is the where I would usually caution against putting too much stock into what the ideology of employees reveal about a corporation’s decision-making, because it is the board of directors that ultimately decides issues of this magnitude. However, Google’s board of directors also appears to be decidedly liberal.”

Here is the description of the data, both under the chart and the authors description of it as it applied to Google, as an example. You and others can go to the site and read for themselves, and you and others can see how the data is compiled.

Here is the Google analysis which has the chart,

The second reason is that, despite its status as a Fortune 500 company, Google is among a growing group of powerful corporations closely aligned with the left. Based on my estimates, Google’s employees are the fourth most liberal of any U.S. corporation, behind Genentech, Apple Inc. and Starbucks (See figure above or click here to view in table format). A table with the To put this in perspective, consider that during the 2008 election cycle, Google employees raised $20,800 for John McCain and $55,451 for Ron Paul; compared to $89,300 for Hillary Clinton and an astounding $803,436 for Barack Obama. In fact, in terms of political contributions, the employees at these firms more closely resemble faculty at liberal universities than traditional Fortune 500 corporations. In some ways, this makes sense. With the exception of Starbucks, each of the most liberal firms, in large part, deals in research and innovation, and each has a reputation of actively recruiting Ph.Ds.


In total, seven members of Google’s board have contribution records that can be used to gauge their ideology. Two of the board’s members, Eric Schmidt (-0.71; view records) and John Doerr (-0.72; view records) are major fundraisers for the Democratic Party "


On the chart labeled "Ideological Ranking of Occupations and Targeted Industries 1990-2008" the author states of the data and association:

"The ideological positions are estimated using an iterative scaling procedure that incorporates PAC and individual contributors that gave to two or more federal candidates between 1990 and 2008"

Notice the word P A C?

You really are just misreading this. Just because the ideological position of PACs is used to help set the ideology scale doesn't mean the data isn't on individual contributions.

Robin, The chart says that PAC contributions are included.

Just because the chart says so evidently in your mind doesn't make it so.

Here in Europe doctors are often left leaning - which may be connected to the fact they are frequently government employees. However, that damages the hypothesis.
Also, the Church of England is full of left-leaning clergy, which may explain is why there is a disconnect on issues such as church marriage for same sex couples and woman bishops between the clergy and congregations.
This again damages the hypothesis.

For the federal government and many states, immigrants and minorities found a more receptive workplace than in the world of business, Employment was either on the basis of "merit" or on the basis of patronage through urban bosses.

Having said that, I think the discussion is too abstract. How about farmers, who gamble on the weather, so are liberal, but have gotten subsidized crop insurance, which is conservative. Or the coal miners of John L. Lewis, who were hardly "liberal"?

Based on the data, which is based on PAC contributions of the employer, Target employees are conservatives.

You're making the distinction here on attitudes to rare events. Attitudes to the payoff of rare events in each career then leads you to typecast on the politics of the profession. This is interesting but surely poorly thought out! For example you've put doctor under conservative, as rare events may pay badly. But perhaps doctors can also benefit from rare events (I identify a miracle cure).

Perhaps a better guess would be to look at how your pay scales with each type of industry. i.e. if I'm paid for each hour worked, then what I earn is limited by the hours I can work, but I receive a guaranteed paycheck each month. Perhaps this might be an indicator of a conservative mentality.

Contra the above, if I decide to participate in an industry where my pay is not directly related to the hours I work, i.e. artist, entrepreneur, I'm perhaps a risk taker as I'm prepared to accept the risk that I might not be paid for every single hour I work in return for a potential payoff based on longer probabilities. Maybe this is an indicator of a liberal attitude?

Verbalists tend to be on the left.

When I was looking into colleges in the mid-1970s, the SAT-Verbal v. SAT-Math differential predicted a lot about ideology on campus. I believe Bennington had the biggest gap then between SAT-V and SAT-M, adjusted for different means and standard deviations.

In a fair and just world these wouldn't be liberal jobs or conservative jobs they would be IMMIGRANT jobs.

All the conservative jobs require disciplined rule-following, while the liberal ones are jobs where one is placed in ambiguous situations that can be interpreted in different ways and where there is no definite procedure.

Civil servant?

Also, union jobs.

Heavy workplace regulation is a thoroughly liberal thing.

(continued) That fits nicely with a point both Hayek and Sowell made, that conservatives tend to trust time-tested rules while liberals have more faith in the power of individual intelligence.

Penalty flag on the play for putting Hayek and Sowell in the same sentence.

I read Conquests and Cultures recently. It ain't Hayek, but it is good.

I had him for Econ in college and he made us buy his book. At that point I vowed never to read a word he writes.

Seems like the "liberal" professions are all hive castes that tend to promote the eusocial ideal and that attract individuals whose descendents likely would enjoy a reproductive advantage under various eusocial regimes. Descendents of individuals in the "conservative" occupations would enjoy better reproductive success in regimes in which "merely social" was adequate and diverse other abilities also confer reproductive advantage.

Male jobs: building, protecting. Also solitary pursuits (coding, long-haul driving), and physical risk.
Female jobs: caretaking, teaching. Social activities (administration, retail), and consensus-building.

"Right" jobs are men's work. "Left" jobs are women's. Hence the poltiical "gender gap."

That's a sizable part of it.

A lot of it is just verbal v. math skills, which correlates to a fair amount with sex.

It would be interesting to check this from before 1965. Did women entering verbalist jobs in large number drive them to the left, or were they already on the left?

The SAT-Verbal v. SAT-Math political distinction was already clearly in place by the time I was applying to college in 1975, so I suspect the roots are deeper than just male v. female, but sex differences are clearly important.

Maybe it's a feminine brain v. masculine brain type thing. Having gone to Rice in Houston, I would imagine that Exxon's women engineers vote Republican more than most groups other than Exxon's men engineers.

It doesn't seem to difficult to do an actual analysis of this. Use some "task content of occupation" data à la Autor, Levy and Mundane (2003) and look if any task types are characteristic for liberal and conservative occupations. Plus education variables.
But I agree with Bill that the ideology data would need to be better than those on contributions by company. How hard is it to survey people to get valid sample sizes at the ISCO 2- og 3-digit level?

The data is contributions by individuals, not by firms.

That is not correct, Robin.

The chart on the association between occupations and ideology clearly states:

“The ideological positions are estimated using an iterative scaling procedure that incorporates PAC and individual contributors that gave to two or more federal candidates between 1990 and 2008″

You can go to the link listed in my comments above.

Robin seems to have a knack for coming up with original ideas that are little bit helpful but are mostly not very good.

The basic categories of identity politics -- race, sex, religion, class, sexual orientation, urban v. rural, etc. -- turn out to be pretty useful at explaining voting. There are some subtle points that aren't well understood widely yet -- such as the crucial role of marriage, affordable family formation, and real estate prices.

Unfortunately, we're not supposed to think very hard about identity politics categories. So, Robin's attempts to come up with new categories based on his SWAGs are very respectable.

They just don't work very well.

There are a lot of voting gaps on these demographic lines, but what's the story that undergirds it? Presumably, people's voting habits differ in these categories for a strong reason. It could just be social herding, true--I'm a big advocate for peer effects. But Robin's idea is that the professional differences are driven by different tolerances for risk and reward. Different aggregate tolerances for risk and reward could ALSO explain those identity politics differences.

You can argue that identity politics drives the professional differences; Robin could argue risk/reward drives identity politics and professional differences.

Are oilmen who drill wildcat wells anti-risk? They sure vote Republican.

I think that's a good counterpoint to Robin's argument. The adrenaline professions tend to skew Republican. You would not expect that if conservatives are more risk-averse (unless I am misreading Robin).

Nascar drivers!

Professional golfers, who don't have guaranteed contracts and have been known to wager a dollar or two out of their own pockets, have been >90% Republican for a long time. Tom Watson was the only touring pro to vote for McGovern in 1972. Scott Simpson might have been about the only pro to vote for Dukakis or Clinton.

Or how about football coaches?

There's an interesting sports v. stories partisan breakdown. People who prefer watching sports to stories on TV are more Republican.

urban v. rural is not a category of 'identity politics' and religious identification per se is generally pretty useless as a variable. IIRC, you need a mess of categorical variables specifying religious affiliation and practice 'ere you find statistically significant results.

Nah, I found the Big One:

Altogether too similar to arguments that more homeowners = more middle class people = more Republicans. There may be a correlation, but the causation (if any) is not obvious, and might work in the opposite direction.

There's a big status difference between conservative and lean conservative jobs (and a smaller status difference between lean liberal and liberal). It appears that people in higher-status professions aggregate closer toward the center.

As someone who has worked in and around newsrooms, academia, and authors, I can say they are very attractive to liberals. First, all three provide a lot of novelty. Journalism puts you in contact with new people, academia and writing new ideas. All three have audiences, as do all the lean liberal professions. They are also dominated by liberal people, and there may be some path dependency at work there.

"All three have audiences"

You may be onto something there. Do liberal jobs tend to garner more applause or public approval than conservative ones? Certainly quite a number of left-leaning professions seem to have partially traded lower pay for higher social rewards.

This may link to why the left seems (at least in recent decades) to judge laws and policies based more on good intentions and less on actual detectable results. Although I admit I may be exhibiting my own bias here.

I think that's a factor.

Being a school teacher is a form of show biz. (If your audience tries to sneak out you can punish them.)

Especially, gay men love an audience.

"Especially, gay men love an audience."

That's hilarious - personal experience?

No. People in many occupations now vote differently than they used to, even though those occupations still involve dealing with rare bad things, or require verbal communication skills, or whatever. For instance, there's a bit of data in the 40th slide of this talk by statistician Andrew Gelman:

Haven't been able to find it just now, but I seem to recall Gelman also showing other data broken down further than in the above link (which shows data for fairly broad categories of occupation). For instance, scientists used to vote Republican back in the 1960s. Now they vote Democratic.

Ok, found the link to further data from Andrew Gelman, but for some reason the figure doesn't display:

Oh well. FWIW, in the post, Gelman notes that various professional occupations (doctors, lawyers, teachers) have shifted from the Republicans to Democrats over time, whereas business owners have gone to the Republicans.

Teachers are unionized public employees, and likely the most crucial component of the Democratic Party vote farm. Also, teacher-training programs are commonly quite politicized and sometimes explicitly screen out people not with the program (see LeMoyne College's for an example). The legal profession and the business sectors are natural enemies much of the time.

The extremely high level of political correctness in teacher training tends to drive out people with half a brain.

Teaching profession has also gone from chemists teaching chemistry to education majors teaching chemistry.

The parties have been stratifying in terms of class markers, especially education, with the Democrats as a Hi-Lo sandwich and the Republicans in the Middle. But that means among white people more people with high status educations vote Democratic. Or at least that's true in the media centers on the coasts.

Scientists by definition have high status educations.

That's interesting. I wonder if there are peer effects here. One thing police, professors, journalists, and authors have in common is that their social circles are mostly full of fellow police, fellow professors, fellow journalists, and fellow authors.

Maybe the strongest effect is that you trend toward what your peers are doing.

If it's a peer effect, it could explain change in the profession over time. Maybe scientists herded around Republicans in the '60s, but then a sea change pushed their herding somewhere else.

Speaking as a scientist myself, I can tell you that in the US and Canada a big reason scientists don't vote for the Republicans (or Conservatives in Canada) these days is because they see those parties as hostile to science. My personal experience only goes back to the 90s. But it's possible that if you went further back, to say the 60s, you'd find a time when conservatives who were seen as science boosters.

I think this is the kind of thing you want to look at if you want to understand why people in particular occupations in particular countries tend to vote a particular way at the present time. Not totally-ungrounded arm-waving about whether certain professions have to deal with big rare losses or whatever.

There do seem to be idiosyncratic factors driving a lot of this. I've heard that Watergate pushed a lot more liberals into the journalism game.

But I'm still rooting for a peer effect. I can't say whether it's big or small, but I've felt the pressure to change my views (I'm a lean conservative who has worked in solely liberal professions). Even the idea that Republicans are anti-science has -some- basis in peer effects, the exchange of memes, and social herding.

I’ve heard that Watergate pushed a lot more liberals into the journalism game.

I believe retrospective survey research indicates that the national press corps was overwhelmingly Democratic by 1964. IIRC, research by Robert Lichter and others undertaken before new media took off indicated that provincial journalists have views closer to population means but that journalists' scores on questionnaires are as a rule to the left of whatever their constituency is. It appears to be an abiding property of the word merchant element everywhere.

Right, the national press was quite liberal in the 1960s. There weren't a lot of conservative sources. The Wall Street Journal wasn't a very big deal yet, and the one Republican paper in New York, the Herald-Tribune, went out of business around 1964.

The three TV networks and PBS were liberal.

The Reader's Digest was a conservative magazine. The Saturday Evening Post had been, but was pretty liberal by the Sixties. Life was moderately liberal. Time was rapidly moving to the left as Henry Luce's influence faded. Fortune and Forbes were kind of conservative while BusinessWeek was kind of liberal.

"I believe retrospective survey research indicates that the national press corps was overwhelmingly Democratic by 1964."

Art: Foolish of me not to check the survey research prior to Watergate. But yes, I've reach Lichter and tend to agree. There's a paper by Gentzkow and Shapiro that also seems to show provincial media skewing slightly to the left of their market (though market effects appear much stronger). It fits my experiences in provincial media, too.

There was that weird early 20th century era when the New York Times was very conservative and pro-business, though. Always curious about that, personally.

they see those parties as hostile to science.

The Republicans might be marginally less inclined to puke federal appropriations into the NIH and the NSF grant money honey pot. Big deal. They 'see' things that way because those are the clothes you wear when working in an academic institution.

You're missing the point entirely. It's not about money.

It's about climate change and evolution, and it's not about vaccines and GMOs and nuclear energy.

I'm not exactly sure what the unifying principal is, but evolution and the fundamentalist Christian position is probably the biggest howler that draws scientists away from the right.

Anyway, it's not about funding. An awful lot of scientists are employed by private industry, or defense money, or things otherwise more attributable to conservatives.

Big difference between global warming oh sorry climate change and evolution.

You’re missing the point entirely. It’s not about money.

Pretty amusing since my point went right over your head.

but evolution and the fundamentalist Christian position is probably the biggest howler that draws scientists away from the right.

What effect do young Earth creationists have on federal policy (or, really, policy anywhere)? Nothing. This is a status marker, and that's it.

"What effect do young Earth creationists have on federal policy (or, really, policy anywhere)? Nothing. This is a status marker, and that’s it."

I pretty much agree. Most of "Republicans are anti-science" comes from hatred of hard-core Christians. Who admittedly, have some pretty laughable beliefs. But it's not like there aren't crazies on the other side.

Christians. Who admittedly, have some pretty laughable beliefs

Sorry, don't think the arts-and-sciences faculty outside the religion department could give you a precis. That aside, faculties happily tolerate the spew of the victimology programs and 'ere that psychoanalysis.

I'm not sure I understand your comment. Young Earth beliefs are silly and their presence in the conservative tent hurts conservatives. Part of the problem is that people extend that to thinking that everything vaguely conservative people say is silly. So people stop listening and "I don't think a large carbon tax in the US makes sense to have a small effect on a 1 degree temperature change in 100 years" gets shouted down with "Climate Denier!" The effects on the gay marriage debate were even more direct - It's hard to find gay marriage advocates who are aware of or have even listened to the arguments against gay marriage, because they associate them with people, religious people, that they hate. It bugs me that this is not symmetric, so anti-vaxers in the liberal camp (who clearly do much more harm) don't get the same attention, but that's the way it is.

Among white people, more people who Don't Believe in Evolution vote Republican than Democrat. So that's a status marker important to scientists. Of course, if you add in all the blacks and Hispanics who Don't Believe In Evolution, the picture looks a lot different, but white Democrats try hard not to think about blacks and Hispanics except in the most symbolic/sentimental terms. For white Democrats, the status game is overwhelmingly white v. white.

Moreover, because Democrats and the media are so closely aligned, they get to decide what is and what isn't science. For example, a whole lot of white Democrats don't believe in evolution from the neck up, but they have almost no conception how anti-science their prejudice is. It just never comes up in their heads.

We live in an era when the reigning prejudices are deeply anti-noticing.

I think authority and religion are the big factors here.

Conservatism is pro-authority almost by definition because a strong hierarchy is the stablest system. While there is an authoritative aspect to science (ie PhDs and pursuit of a canonical truth) in practice it's very much about disrupting old beliefs with new discoveries. This will draw more liberal thinkers. As a consequence I'd predict science to be more liberal than engineering because engineering is much more about optimizing the existing order.

But the bigger factor is religion. The religious faiths simply don't stand up that well under science, religion is generally conservative (I'd argue due to authoritative aspects), and so science becomes an enemy of religion and therefore conservatism.

Again one prediction of this is that science would become much more liberal around the same time that religion started to feel under threat by secular and atheist arguments.

If you look at voting by state in Presidential elections, the main drivers in the 21st Century -- Years Married among Younger White Women, the Dirt Gap, Affordable Family Formation, etc. -- are pretty novel. Partly that's the impact of Perot running in 1992 and 1996. If you go back to 1988 for a 2 candidate election, you can vaguely see these patterns starting to emerge, but they are nowhere near as stark as in the four elections of the 21st Century. The correlation between Year's Married and Red v. Blue over the last four Presidential elections is just crazy high:

Does the marriage variable do a lot to explain these professions? I actually think professors have a high rate of being white and married. Journalism and writers too (though most authors I know are women--the stereotypical author in my experience is a married woman who is also a homemaker or works part time).

Marriage might explain aggregate trends, but how well does it perform for the sorting in these occupations?

Age of first marriage is one key. Professions where people tend to get married late, like academia, tend to be liberal.

One basic bit of logic is that the Republicans position themselves as the Family Values party, so the earlier in life people have legitimate families, the more likely they are to decide they belong to the Family Values party.

But the arrows of causality run in multiple directions. People who want to get married and have kids might move to Red States with cheap houses. But people who want to move to San Francisco find they can't afford to have kids. Or people born in San Francisco absorb the surrounding culture.

I'd like to see studies done of sisters or even twin sisters who get job offers out of college in different places and see how their politics evolve.

There is one political scientist currently who is mining my ideas from 2004-2005 and getting a lot of papers out of them, but there's a lot more that could be done to research these patterns. Economists could easily do the analyses.

Political discourse would be much better informed if Nate Silver had invented Affordable Family Formation. Your brand hurts a really clever idea.

If you ever get another one this good, do us all a favor: Sneak into one of Tyler's friend's offices and scribble something close to it on the whiteboard. :)

I think Affordable Family Formation is a *very* clever idea that explains a lot about our politics. The next decade as the rustbelt Midwest shifts rightward and the expensive liberal coasts shift further leftward will prove it beyond a doubt.

Of course I wouldn't have come up with it if I weren't such a horrible person and inveterate crimethinker. It's basically a rediscovery of the proto-Malthusianism that Ben Franklin came up with in the early 1750s that helped get Darwinism off the ground a century later. That's the central thread in Anglo-American intellectual life over the last quarter of a millenium, but it's lost today.

It would be interesting to know how audiologists and speech and language pathologists lean. They are surely less conservative than dentists, more conservative than psychiatrists, maybe.

You can find that out in the link I posted above, because there is a piece discussing the ideological political contributions by surgeons, doctors, etc. What was interesting was that as doctors became employees of hospitals they became more liberal based on contribution changes for the group.

Creativity and comfort with ambiguity are traits that have been shown to correlate with liberalism.
There's also likely a geography-related hidden variable lurking here. Urban vs. rural sorting in where jobs are clustered.

It's also useful to consider data from outside the US. If some jobs really are "conservative" or "liberal", then they should be conservative or liberal everywhere. They're not. See, for instance, these data from Australia, where people in various jobs Hanson calls "conservative" (e.g., various construction and heavy machinery operation jobs) vote Labour:

I think there's a distinction between salaried and wage employees in these circumstances. I suspect if you look closer, political splits among the working class in the U.S. tend to be delineated by personal status, residence and ascribed characteristics (i.e. married v. unmarried, metropolitan v. non-metropolitan, South v. the rest, men v. women, whites v. blacks v. hispanics); political splits among the bourgeoisie would be influenced by occupational type with some influence from personal status (married v. unmarried) and cultural affiliations (South v. the rest; church attendance).

The cultural factors do not necessarily translate well abroad.

Jeremy, You are right, but also the data is screwed up unless you believe that PAC contributions should be attributed to employees working in that industry. The problem with this post is that the data collects PAC and director contributions and attributes them to employees.

Of course you wouldn't find the correlation. The world is more complex.

"Religious worker" is not a 'conservative' occupation. It might be (if not universally so) among evangelicals, pentacostalists, Mormons, the minority Lutheran sects, a small corps of traditionalist Catholics, and the Eastern churches. Mainline protestant clergy are liberal by default, and commonly complacent and fuzzy-minded for 'a that. The personnel of the Catholic Church in the United States clerical and lay consists largely of tribal Democrats; many of them have no time for the positions of the Church which conflict with the zeitgeist and most of the remainder are assiduous if not talented at attitudinizing and excuse-mongering. Traditionalist Catholics often have an affinity for weird alt-right politics and have a truncated interest in civic affairs. The politics of Eastern clergy are often very opaque. Evangelical laymen have been discovering in recent years that their academic institutions are run by people who are embarrassed by them and despise them (see Randall Ballmer) and are now discovering that evangelical civic apostolates are run by geldings (see Russell Moore for an example, or the losers who have succeeded James Dobson at Focus on the Family).

Today I learned that, via Tyler and Robin (despite the disclaimers of the author disclaiming extending PAC contributions to political views of employees) that

Target employees are conservatives,


Starbucks employees are liberal.

Bill, do you honestly doubt that Target employees are conservative and Starbucks employees are liberal?

Lord, Lord, Lord,

You missed the joke.

The data on this study attributes corporate PAC contributions and director contributions to the companies employees. Target (PAC and directors) contributions are attributed to the employees; same in the other direction for Starbucks.

Which is all to say that the whole story of the post is BS since the data is BS. Go to the link in my initial comments and look at the chart in the Google story, the heading, and the description of the data and how it is aligned with occupations.

You moved from saying it _was_ PAC data to saying it _included_ PAC data, and you haven't told us why that's a serious source of bias. It doesn't seem to be. I'd expect Target employees to be conservative and Starbucks employees to be liberal, for example. Wouldn't you?

I think Hanson is ignoring a lot of obvious explanations, but you vomited a weak criticism all over this thread and acted like it was the most important thing for everybody to notice. One comment would have been enough.

Cmon, I said it includes PAC contributions. The PAC and board contributions determine whether the EMPLOYEE is liberal or conservative based on the charts and the description of the data.

You can characterize pejoratively my criticism any way want, but evidently you are unable to deny it.

I cannot help you with reasoning.

When I worked for D&B 20 years ago, I was informed that it was customary for executives to pony up something like 0.5% of their salaries to D&B's PAC. When my wife worked for Cook County as a computer programmer, it was made clear that she was expected to donate a larger percentage of her salary than that to the re-election fund of, I believe, the County Treasurer, who was the head of her division. (I vaguely recall that lady eventually being put on trial for campaign corruption by the feds, but I could be wrong.)

I'm not sure if the same occurs in the US, but in Canada there have been a couple disconnects. Large unionized firms such as auto manufacturing would have conservative management, socialist union leadership and neither would predict how the rank and file would vote.

In BC the Left lived off the working class people, the miners, lumber workers, construction workers. The unions were strong and had deep support, and their members voted reliably NDP. That has changed in recent decades, and the change is somewhat country wide. The NDP and the center left federal Liberals now represent the cities, the government workers, the socially progressives, the environmentalists. The workers are conservative because they find that the other parties want to put them out of work, and in fact have put them out of work. Alberta is now staunchly conservative and filled with what formerly would have been NDP voters; skilled trades working for large corporations. But the NDP wants to shut them down, so they don't get the vote. They are unionized, or a substantial percentage of them are, they are paid very well and the conservative provincial government spends per capita more on social spending than any other province, but the ideological left hates them. Oil.

Something I've found in Michigan is a lot of old UAW guys, gun owners, severe social conservatives, many of them vietnam vets whose every political impulse is conservative. And every fall, they vote a straight democratic ticket. Old habits die hard.

I'm not surprised about police, religious worker, and insurance broker being conservative. But doctor? I'm always hearing claims of conservatives being "anti-science," so I guess I thought MDs were involved in making those claims. Liberal propaganda worked on me I guess.

About 80% of scientists I know are pretty conservative. It's only the left that paints the “anti-science” picture.

Dwelling on this subject is doing nothing but making us a more partyist society.

We are sorta stuck with parties because a couple of centuries of experience tell us that they are a part of our political equilibrium. The founders eschewed "factions" as they were called, but the ink wasnt dry in their thoughts before factions formed.

The biggest donors, at least correlate with subsidies special tax treatment Agriculture (crop supports, ethanol mandate), Oil and Gas, Mining (percentage depletion), Banking (too big to fail). They realize that Conservative rhetoric is a screen for less spent on unworthy causes that shift income to and reduce taxes on low and middle income people, leaving more for them.

So, Billy, what did you learn in school today at the MR website using the data they gave us associating corporate PAC contributions of the employer with the employee.

Well, I learned that

Tobacco workers, defense workers, truckers, coal miners, insurance agents, residential construction workers, auto dealers and gasoline station attendants are conservatives

and that

Venture capitalists, hedge fund workers, teachers, psychologists and psychiatrists, corporate lawyers, professors, and homemakers are liberal

That Target workers are conservative and that Starbucks workers are liberal.

This strikes me as very silly. The methodology for determining what constitutes a liberal or conservative job is pretty bad to start with. Then only a few jobs from each category are selected 'after the fact' to fit the narrative they support. What's more many of the jobs used to support the narrative barely lean one way or another. Those jobs that lean strongly toward one party or another all have pretty direct, if idiosyncratic, reasons for supporting a party. Farmers receive direct subsidies that are directly supported by (mostly republicans) energy extraction industries would be hurt by regulations that are opposed by republicans etc. What's more such direct benefits are usually an emergent property of messy coalition politics, IE farmers are more likely to receive subsidies from republicans because they support republicans rather than support republicans because they receive subsidies, indeed in places where democrats lead a coalition that includes farmers they support agricultural subsidies.

I think these sorts of thought experiments can be valuable, but holy cow are the conclusions here weak.

A better methodology would be direct surveys of workers in a field on how they voted in the last Presidential election. For example, golf magazines intermittently try to find a Tour pro who votes Democratic, and it always takes some legwork to find one.

I'd guess someone else has said this, but I think he might have the causality the wrong way. Teachers are liberal because they think liberals are more likely to give them more money, for example.

Just as defense contractors are liberal because they think that government will give them money too.

Turn on the lava lamp, play "The Dark Side of the Moon", and read the Old Right critiques of Wilson's and Roosevelt's defense spending. It'll blow your mind.

I'll lend you the lava lamp so you curl up with an Ayn Rand novel. The point is that "causality" applies both ways, but the commenter didn't see that or chose to comment on it.

How much of the federal budget is spent on defense, how many wars were fought where we didn't raise taxes to finance them but simply ran up a deficit to pay for them.

Why would they think that?

It's all about the promotion pathway.

Liberals tend to be empathetic people. Liberal occupations are ones where success and promotion is caused by the mastery of the ambiguous and shifting social environment.

Conservatives tend to be principled people. Conservative occupations are ones where success and promotion is caused by the mastery of something with unchanging underlying principles.

Okay, but a lot of Republicans sell and buy stuff -- stocks, real estate, etc. -- where the price changes constantly and noticing trends is important.

You say that liberals love risk, consvertives avoid it. How do reconcile this with the fact that men tend to be risk takers and conversative, but women tend to be risk-averse and liberal?

When certain people gain a stranglehold over admission to their exclusive club, the perceived effects of self-selection begin to dwindle.

Police officers may be socially conservative but they and their labor union tilt toward the liberal politicians who hand them taxpayer dollars.

I see no inconsistency with the theory that conservatives tend to minimize maximum regret. It is almost tautological. The risk loving nature of liberals appears to explain a lot of monumental successes and failures, but I am unimpressed by successes built upon risks not borne by the policymakers themselves.

Are tenured college professors risk loving relative to wildcat oil well drillers?

Well. Anyone pursuing a tenure-track job in the Humanities right now clearly loves risk a little too much.

Or not perceiving the risks, or avoiding the risks associated with abandoning academe.

One needs to look more carefully at the dispersion also. Note that although physicians are "conservative" on average, more than 40% of them are liberal. Other professions like investment bankers, also have a huge variation.

I think there is something to Robin Hanson’s theory that “Conservatives are more focused on fear of bad things, and protecting against them.” Maybe it is more accurate to describe it as a “Conservative way of thinking” in which both Conservatives and Liberals partake. I own a small moving company and frequently drive a moving truck on jobs. I’m aware that the most important thing in that job is to not crash the truck! The moving job may go well, and I may make decent money on it, but the downside of a truck crash is always much greater than the upside of that job’s expected profit. In other work activities like an actor auditioning for a big role, the upside seems much larger than the downside of simply not getting that particular role.

The pattern really seems to be:
1) Those focused on with materials -- cars, autos, and minerals tend to be conservative.
2) Those who have smaller business that employ people who do physical work are conservative (e.g. construction, mining). Those people probably have a lot more frustration dealing with regulation, as dealing with regulation often requires higher earning talent that does not directly produce the product.
3) People who work primarily with ideas tend to be much more liberal; except for those in high paying fields like investment banking and lobbying. The latter have more diverse views.
4) Physicians are diverse because they include small businesses, people who want to earn a lot of money, and idealists.

Jon, The data is based on PAC contributions. So, if you are a tobacco company, and contribute to conservative candidates, ergo, your employees are conservative. See my comment on the data above.

What about Software Engineers? Mostly liberal, but don't fall into either category.

Actually, according to the cooked data, software engineers at Apple and Google are very liberal. Of course, you get this way because the data Tyler cites attributes director and PAC money to the employees. Oila.

This makes me want to create a liberal PAC for George Mason's economics department so I can attribute the contributions to its faculty members.

Perhaps empathy is the key variable. Doctors, cops, insurance brokers often don't view their customers as individuals. A doctor has to face death often without letting it get to him. Cops must view criminals as criminals without too much *understanding* getting in the way.

Teachers and psychiatrists are high on empathy. Lawyers must have a good sense of how others feel about things, just like journalists and artists.

So the conservative group is more objectively analytic while the liberal group is more touchy-feely, which corresponds exactly to stereotype.

psychiatrists are high on empathy.

Do you know many?

I'm in psychiatry residency. Anecdotally, the split seems fifty fifty

Just another variation of the ad hominem argument.

Hey, Rich, this isn't ad hominem. If you believe the argument, if you are liberal, you are with professors, hedge fund owners, Apple, Intel and Google engineers, creatives, corporate lawyers, venture capitalists, psychologists and psychiatrists, and homemakers, and conservatives are truck drivers, insurance agents, coal mines, residential construction workers, and gasoline service station attendants.

Of course, if you look at the data, you would see that what they did was attribute political contributions of the PAC and directors to the employees when they made this assertion.

If I work for Exxon, therefore I am a conservative, even though Exxon employs many types of employees, probably even corporate lawyers, psychiatrists, software engineers, creatives, etc.

For those saying that a professor's job performance is not measurable....ever had to worry about obtaining 5 "A" publications (with reasonable citation counts) in order to hang on to your job?

I would argue that the average white collar cube jockey's productivity at pushing papers around all day is much harder to measure. This is coming from an academic liberal of course...

Seems like the conservative jobs tend to be more literal, pragmatic, direct, and concrete. The 'liberal' jobs tend to be more abstract, intuitive, hypothetical, and mental. Note that the religious workers might be more easily viewed as abstract and hypothetical rather than literal and concrete; however, if you assume that most of the conservative religious workers are approaching their work from a traditionalist, rules-are-rules, obey-your-duty-to-God perspective, then it plausibly fits alongside jobs like doctor and building manager.

Alternatively, the conservative jobs focus on more specific rules and structure, while the 'liberal' jobs include lots of wiggle room within broader principles.

What is "conservative" and what is "liberal"? There are papers cited various places, probably included here, that note very low correlation on views on various topics. So your "liberal" lawyer might favor government spending to improve education, be for gay marriage, be against many regulations, be for an orderly society and thus have some degree of law & order view, etc. What does it mean to say "liberal"?

Do we really mean "favors state and quasi-state employment is a huge part of the economy, maybe all of it" for liberal? Do we mean "respect for property rights and responsible behavoir" when we say conservative?

Charles De Gaulle favored a massive role for the French state in the economy. It's hard to call him not a conservative.

In France we don't devide people between conservative and liberal. Liberal is used both for left leaning or right leaning people.
Division is on 2 axis. One goes from progressist to conservative, the other from liberal to statist.
De Gaulle was a conservative statist. Sarkozy was a conservative half liberal. There are almost no real liberals in France.

Some of the distinctions between industries are just WASP v. Jewish. Oil is mostly gentile so it's Republican. Hollywood is Jewish-dominated so it's Democrat.

Here's a thought-experiment. If Steven Spielberg had been adopted by a gentile family in Arizona and went into, say, real estate development, would he incline toward the Democrats or the Republicans?

My guess is that he's naturally conservative, a loyalist. In this universe, he's loyal to his Jewish ethnicity and to his industry's traditions, so that makes him a major Democratic donor. But, it's not hard to imagine another universe in which he's a brilliant Scottsdale developer and pillar of the local Republican Party. (There's a curb your enthusiasm episode in which Larry David temporarily learns he's really a gentile from Arizona.)

Again, Hollywood engages in the manipulation of words and images.

A better example might be community banking v. casino banking.

I have to use theological terminology that nobody here would probably understand. A good part of it comes down to your position on dogmatic law. The closest thing you might get in secular terms would be authority. Conservative fields have no problem and in fact prefer specific laws. They don't necessarily have to make sense. Sales guys have to deal with the more or less arbitrary quota. You did good last year, ok, to make the cruise this year, sell 20% more. Conservatives are comfortable with axioms. Hence the tilt as mentioned above to Mathematical ability. Liberal jobs in contrast are those where the rules keep shifting and if you can talk well enough you can game the system to outsize rewards. The pastoral jobs are the canary in the coal mine. The conservative leaning clergy are those who preach both law and gospel. The mainliners (i.e. liberals) major in subverting the law or explaining why it doesn't mean what the black letters say.

The axis of importance as I would see it is

Liberal jobs are ones where it pays a lot to be able to think carefully a lot about abstract ideas and infer other, unfamiliar mental states (Hanson would probably call this "far")...

... but where there is a relatively low premium on self control and self regulation.

Empathetic, imaginative, offensive, uncontrolled Liberals vs polite, restrained, conventional, unempathetic Conservatives.

Equally, jobs where it is a real disadvantage to spend your time fixating on abstract ideas and mental states, yet where at the same time it's important to control yourself, show focus and determination and not "wander" will tend to lean Conservative (in our knowledge>self control bailout economy and society, these often tend to be fairly low prestige jobs, but not always).

Liberals tend to be people who like fairly abstract ideology combined with the fact that this ideology encourages and rewards low degrees of self control and regulation (whether its libertine or paternalistic).

Conservatives tend to be people who are rather self controlled, but fail repeatedly, eternally and completely to envision and grasp ideas outside their experience which undermine their boosting of Doozer like individual striving and "personal responsibility". So most often are perennially these scammed and exploited stooges for the people who understand how things really work.

Libertarians are broadly like Liberals, but tend to be rather more affectless (either on an autistic or psychopathic dimension), so lack much of the empathetic imagination of Liberals.

On the other hand, there aren't a lot of jobs where these are markedly different, so the real terms difference for most jobs is probably not so high and political factors such as economic support from government inflate the apparent differences.

My initial take:

Liberal = knowledge + teaching/informing others + creativity, Conservative = $ + doing stuff

Actually, not. Read Haidt's book listed in the comment below. More complicated. Here is a summary link:


Rather than speculate, why not read Jonathan Haidt's book: The Righteous Mind: Why Good People Are Divided by Politics and Religion (Pantheon 2012), a professor of Psychology at the University of Virginia. Take the Chicken test and the other tests, including empathy, self and other control, fairness, loyalty/cheating, authority/subversion, and sanctity/degradation, and the care/harm tests. I think you will like the book.

tmc, no. that's not true. read the links and the data description in table listed in the links below.

I really think upbringing is being overlooked here.

It seems to me that a liberal upbringing (like the one I received) will tend to interpret a lot of behaviours the parents cannot account for as stemming from deep personality traits. This style of parenting will tend to double down on ('celebrate') what a child is good at, and resist pressuring the child to master tasks and skills that seem to run against the grain. The parents' bias will be to test a wide range of extracurricular activities in order to find a good 'fit' for the child, something they naturally enjoy. They will encourage the child to think similarly of professions later on.

A conservative upbringing, on the other hand (like the one my fiancee received), will tend to treat personality as more malleable; conservative parents will treat personality mostly as a collection of habits and focus on the formation of 'good' habits through practice. Thus they will put pressure on children to perform well in 'important' subjects at school and to take on extracurricular activities associated with 'successful' people or character-building. Similarly with professions later on.

The problem of course is that a) many professions rely on specific motor, quantitative, analytical, spatial, or language skills that are best acquired at an early age; that early advantage cannot be easily overcome later (perhaps it cannot at all). and b) learning and working against the grain is harder, less satisfying, and possibly less effective.

Hence the liberal upbringing will tend to shut children out of careers on the basis of relatively early-life preferences, possibly very superficial ones; the conservative upbringing will tend to force children into careers on the basis of parental biases, and deliver a large cohort of hard-working but less productive and less satisfied people.

I don't know that there is such a thing as a 'libertarian' upbringing, by the way. I think parenting styles are on a continuum of the above two; on the other hand, whether or not children are taught to share, and/or to expect privacy at home, could influence their views on cross-subsidies later on in life.

Do you have some evidence regarding the political-affiliation-based parenting habits you posit? Color me skeptical. There probably are systematic differences, but I wonder whether they are what you think they are. Lots of left-leaning parents put pressure on their children to succeed. Lots of right-leaning parents (particularly religious ones) value traditional measures of success (such as getting a high-paying job) less than their peers.

No more evidence than most opinions on this blog I'm afraid. Moreover, even if I'm right about parenting styles, there's no guarantee they will map 100% against either parents' or children's politics.

And of course 'liberal v. conservative' doesn't correspond to 'left v. right.'

That said, if you do want evidence I can try to dig into the European Values study (of 2008 I'm afraid, as it takes place every 7 years), and try to test some of this. Weighted representative sample of 65k people I think.

I should also note no one else has been asked for evidence so far; perhaps you'll start a trend?

Not many people are making interesting, novel claims.

Here you go - parenting styles v. two tests of political leaning. The source is, once again, the European Values Study of 2008.

And a spreadsheet with the occupation x political leaning data in Europe (plus Russia and the Caucasus).

Key problem is that it aggregates enormously different political systems and cultures. Left-leaning in the UK is not quite the same as left-leaning in France, and Right-leaning in Germany isn't the same as Right-leaning in Russia.

Another way of looking at it would be to say that conservatives tend to cluster in jobs whose primary products are tangible while liberals tend to cluster in jobs whose primary products are BS ,or to put it more politically correctly,ideas,opinions,conjectures,emotions,or feelings .

I think that for some professions, it's a cultural construct. For instance, the military should be a pretty even split, it's drawn from around the country, there are disproportionate minorities etc. But it isn't. And from my experience, the vast majority of "boot" privates don't have any sort of political inclinations at all. But after a few years of hearing your profession cheered by one side of the aisle (sometimes embarrassingly so) and denigrated by the other side........well, you see all those guys who didn't care lining up Republican.

There are some input demographics and inclinations that help, but the perception and reality that Left = anti-military sentiment is enough to turn some percentage of military folks right. Of course, the left will protest that they love soldiers! They just disagree with certain foreign policy issues. And that's fair, but that's not how it plays out. When soldiers are jammed up, it's conservatives who raise a stink. Having a combat tour or three will get you a long way with the ladies in Houston, and a long chat with the cops in San Fran. Wearing a Purple Heart cap to an Ani DiFranco concert will get you three hours of lots of personal space and death stares, wearing it to a country concert will get you mobbed with folks shaking your hand. And I love Ani and hate country. But I know who is on my side and who isn't, as easily as any child does at school.

Morality: more legislation around morality
Economics: the individual is smarter than the group + economic externalities are mostly myth therefore limit legislation around externalities

Morality: more individual choice around morality
Economics: the individual is not always smarter than the group + economic externalities are non-trivial therefore responsible legislation should exist around externalities

Lawyers and clergy both have great internal diversity of ideology by subtype (specialty or denomination) and basically track the ideology of those whom they serve.

I'm pretty sure that 1 in 3 teachers have been republican, with another 1 in 4 or 5 unaffiliated. A little under half are Democrats. That's been consistent for 30 years or so. Here's one cite: but I've seen it many places.

As a Republican who suffered in ed school and lives in a deep dark blue region of a very liberal state, I'd observe that Dem teachers aren't all the loopy liberal progressives. Most English and history teachers definitely fall into that category, but a lot of the rest are more blue collar pro-union solidarity Dems than bleeding hearts.

"The extremely high level of political correctness in teacher training tends to drive out people with half a brain."

I'd remind you teachers aren't as stupid as most people think--which means a lot of them go into ed school knowing they have to fake it. Which is not to say I don't agree that teacher training expects a huge degree of political correctness. In fact, I think if ed school could drop that expectation, it'd be about three quarters fixed.

Ironically, "reformer ed school" requires far more dogma and conformity than progressive ed school--and it's just as politically correct. Weird, huh?

"Being a school teacher is a form of show biz. (If your audience tries to sneak out you can punish them.)"

Better than that: we get to boot hecklers.

Comments for this post are closed