Political Science and Political Journalism
Via Henry Farrell, Matt Bai comments briefly on political journalists’ view of political scientists:
Generally speaking, political writers don’t think so much of political scientists, either, mostly because anyone who has ever actually worked in or covered politics can tell you that, whatever else it may be, a science isn’t one of them. Politics is, after all, the business of humans attempting to triumph over their own disorder, insecurity, competitiveness, arrogance, and infidelity; make all the equations you want, but a lot of politics is simply tactile and visual, rather than empirical. My dinnertime conversation with three Iowans may not add up to a reliable portrait of the national consensus, but it’s often more illuminating than the dissertations of academics whose idea of seeing America is a trip to the local Bed, Bath & Beyond.
I think this kind of attitude is not universally shared, and generally leads to bad political journalism. I think it’s obvious to anyone who thinks about it that the features of journalism—original reporting, first-hand conversations, speed, granularity—allow it to push the frontiers of our understanding beyond what rigorous political science could possible do. At the same time, it’s just incredibly foolish to go about doing the work of journalism about politics devoid of any broader theoretical or empirical foundations provided by political science.
The events of the day play out against a larger structural backdrop. And it’s just not possible to try to understand them a-theoretically. What journalists unschooled in political science tend to do is to substitute prejudice for understanding. So you notice that in Maryland and Virginia there are a lot of well-to-do Democrats and start writing stories which presuppose that poor people are generally Republicans and rich people are generally Democrats. An alternative approach would be to read Andrew Gelman’s book and you’d see that this is an idiosyncratic feature of a small portion of the country and that, overall, high income is a strong predictor of Republican voting.
Reading Gelman’s book isn’t a substitute for interviewing people or trying to understand campaign strategies. But it provides you with an accurate understanding of the larger context in which to situate those interviews. If you don’t read it, you won’t understand your reporting properly.
45 Responses to “Political Science and Political Journalism”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
March 12th, 2009 at 8:48 am
Matt Bai is an idiot, and should be digging ditches instead of being overpaid to pretend to do political reporting (a task for which he is extensively unqualified). His dinnertime conversations with three Iowans are revealing of nothing except Matt Bai’s prejudices.
March 12th, 2009 at 8:56 am
I am a political scientist and there are times that I enjoy reading Matt Bai, and journalism does have a lot to offer, but it is so disappointing to read this statement, which is a basic misunderstanding of what political scientists do – framing hypotheses and systematic collection of data. His reaction reminds me of all the old time baseball writers screaming about statistical analysis of baseball. Whether Bai or the baseball writers want to admit it, systematic analyses of political science does impact how he views politics and they view baseball. Look at the power of OPS for baseball analysis, or how we accept that judges are ideological (more from my field)
March 12th, 2009 at 9:11 am
I may not understand all this weirdo “gravity” speak from these isolated physics freaks, but I can get a lot more ‘illuminated’ on the subject by talking to 3 people whose lives were changed when they fell down.
March 12th, 2009 at 9:19 am
anyone who goes on about how they learn more from “talking to regular people” than from studying or reading dissertations is to be ignored. haven’t we had enough of this mentality by now? why even take it seriously anymore. these are not serious people.
March 12th, 2009 at 9:24 am
“I’m a journamalist — I don’t need no damn book larnin’.”
March 12th, 2009 at 9:24 am
Shorter Matt Yglesias: Journalists should learn the difference between the individualistic and ecological fallacies.
March 12th, 2009 at 9:26 am
The difference between a journalist and a political scientist is that one is happy getting three people to tell him what he already knows and the other needs to spend six years and thousands of dollars.
More seriously, academics are often consumed by questions that no journalist would ever take seriously. For example, asking to what degree constituent public opinion influences congressional voting can be a fruitful dissertation, but would make a miserable article in the New York Times.
And, if Mr. Brai would have bothered to even interview a single political scientist, he might have found that the “scientist” in that term is rather more aspirational than descriptive. But hey, why ruin a perfectly good generalization.
March 12th, 2009 at 9:27 am
I hate to repeat myself, but the NYT doesn’t hire just mediocrities, but manages to find, again and again, *quintessential* mediocrities. They obviously think it’s their job to do so.
Bai’s statement is willfully stupid, and it’s not his first one. Pretty elementary mistake: An and/and is transformed into an either/or with a mere mention of a retail store. This country really deserves better.
March 12th, 2009 at 9:29 am
Are there more obnoxious people than political journalists?
March 12th, 2009 at 9:32 am
They’re just channeling the spirit of the late Nebraska Senator Roman Hruska, who in an attempt to defend one of Nixon’s dreadful Supreme Court nominations opined that mediocre people deserved to be represented on the Court.
March 12th, 2009 at 9:36 am
Political Journalists, like people generally, think in narrative. Here Bai is basically celebrating this. The problem is if they aren’t reading studies that actually use empirical methods to determine, say, whether religion and ballot initiatves were the deciding factor in 2004, then they can only accept and propagate a narrative that “feels” true. If something sounds plausible and nicely explains the latest trend then it will win out in the “battle of ideas” among political journalists and become reality. This leads to nonsensical meta-shouting on cable and the odd acceptance of journalists that there job is to analyze the “message” of what politicians are saying instead of analyzing whether the “message” is realistic or plausible, forgetting that just because something feels true doesn’t mean it is and that cynicism isn’t skepticism.
March 12th, 2009 at 9:43 am
If I were a journalist, I would be committed to availing myself of whatever sources of information and analysis were necessary to follow, understand, complete, and tell a story.
Those would include interviews, oral histories, documentary and electronic evidence, and, yes, whenever appropriate and helpful, scholarly research.
And as much skepticism as I muster about the general activities and foci of the social sciences, it’s laughable that mainstream journalists would dare mock academic researchers for their lack of insight. Really.
Hell, undergrad students face much tighter restrictions on the quality of evidence and arguments allowed in their term papers than do major reporters when it’s time to print whatever nonsense and lies their editors think is fashionable to print.
March 12th, 2009 at 9:48 am
David Leonhardt is someone who writes great analysis using research, following the news, and even some anecdotes that illuminate. But he does this for economics. We need some David Leonhardts in realm of political journalism.
March 12th, 2009 at 9:53 am
14: The amazing thing is that many journalists do actual hard work. It is so remarkable that we have a profession where the entry level job is so much work, but once you get to a certain level you can basically show open contempt for that kind of research.
March 12th, 2009 at 9:58 am
Shorter Matt Bai: Derek Jeter is so a great defensive shorstop! Who are you going to believe, eggheads with their data or me with my first hand account of 3 Yankee games?
March 12th, 2009 at 10:10 am
DTM: Well, not only would it be hard work, I’d be massively undermining my market potential as an influential, behind-the-scenes-small-talk propaganda hack, so I guess that’s why I’m not a journalist. In fact, it’s why I had to spend so many hours and years of my life helping out local alternative press.
March 12th, 2009 at 10:30 am
The idea that a college professor in Columbus, Ohio or Lawrence, Kansas is farther removed from the life of everyday people than a New York Times journalist is pretty funny. And when I read the phrase “whose idea of seeing America is a trip to the local Bed, Bath & Beyond”, I thought he was talking about David Brooks.
March 12th, 2009 at 10:38 am
The fact that people (sorry, “elites”) like Matt Bai exist is seriously depressing. The system is dysfunctional.
March 12th, 2009 at 10:41 am
He is griping about stats (so yes, this is along the lines of the old-line baseball writers not liking Bill James). Real stats, in this case, would require some knowledge of mean vs median, standard deviation, standard error, and such. There are also the issues regarding sampling (the “dinnertime conversation with three Iowans” is what would be called a “convenience sample” and while it gives fun stories, is more nearly comical in terms of analysis). Some newspapers have employed (basically) statisticians to go through the stats and look for patterns – I believe one of the Miami papers had someone who did that and who went on to teach some of this at a journalism school. But most reporters have no understanding of it; and it does not easily lend itself to good stories. Bai shouldn’t really brag about this, as it amounts to saying that he doesn’t need a dictionary (or spell-check) because his understanding of the correct way to spell words is best (uh, Matt???). But I don’t see this changing. Even though vote predictions (based on sampling) are usually quite good.
March 12th, 2009 at 10:47 am
Love that line: What journalists unschooled in political science tend to do is to substitute prejudice for understanding.
That’s the difference between Jim Lehrer/Gwen Ifill and David Brooks.
March 12th, 2009 at 11:10 am
We need a serious study of the social skills exhibited by political scientists at the Applebee’s salad bar. We may have to build some Applebee’s salad bars in order to do that, but once we do, we can watch and see whether or not the political scientists act all dorky at the salad bar or if they are like a real cool salad bar brat pack, just like insider pundits and journalists would be.
March 12th, 2009 at 11:14 am
As a (former) academic political scientist I am surprised to see so much uncritical support for the value of PS. The discipline has access to very powerful tools, and some sub-fields (public opinion, for example) are very fertile and generally recognized outside the academy. However, I am struck by the overall lack of visibility of the discipline in public debate. I tend to attribute this to the habit within the discipline of asking trivial, academic questions, and obtaining, as a result, insignificant and irrelevant results.
March 12th, 2009 at 11:36 am
Roland Stephen: I assure you that to whatever degree I award respect to political scientists, it is not uncritical.
March 12th, 2009 at 12:33 pm
If Bai were better exposed to the field of political science, he might have come across the work of Shanto Iyengar at Stanford. His experiments have shown, among other things, that when media reports use personalistic frames (ie, the story of a family coping with unemployment) versus broader social frames (ie, aggregate unemployment trends, proposed causes, macro-economic impact), audiences are more apt and able to form policy-relevant responses in the latter case than in the former. In short, when the media focus on “dinnertime conversation with three Iowans,” their audience is no better equipped to participate in the democratic process than had they watched the latest episode of “The Biggest Loser.”
March 12th, 2009 at 1:07 pm
You do realize, of course, that BOTH fields suck.
March 12th, 2009 at 1:11 pm
I’m actually pretty sympathetic to Bai’s statement, but then my issue with “political science” is actually a lot more fundamental:
That the very idea of a science of human behavior requires said human behavior to be dictated by forces that can be scientifically understood; to be scientifically understood, a force must be consistently predictable; to be consistently predictable, that force must “be exactly” — that is to say, it is limited by what it is, and thus defined by its limitation; what Sarte referred to as “being-in-itself”.
Such a force, so defined, by definition cannot be free; thus free will, and the behaviors it dictates, cannot be scientifically understood.
This is how I understand it; so, unless I’m wrong, a study of human behavior, to the degree that it seeks to use a scientific model to study it, makes the implicit claim that said behavior is not subject to free will, if free will can then be said to exist at all.
March 12th, 2009 at 1:12 pm
“If people do differ from rocks in this respect, then a science of human behavior will have to take account of this fact.”
Noam Chomsky
March 12th, 2009 at 1:45 pm
@El Cid: Let’s at least compare the best practices of political scientists and political journalists — not the best political science with the worst political journalism. While they tend to happen before a smaller audience, the practices of lazy or cynical political scientists can be no less jaw-dropping.
One book in particular made me think twice about my impatience with certain journalists. It’s by the often profane Joe Bageant, and the book is often personal and anecdotal, the kind of stuff that often tries my patience. Yet Bageant also did his research, and foretold the mortgage crisis and so-called “meltdown” in 2006, from a perspective that few journalists and no political scientists, to my knowledge, have assumed. Briefly, Bageant describes the incentives of the mobile home dweller to move up to a mortgaged prefab home, “no matter what it takes”; of the mortgage lender to finance the deal on terms far in excess of the property’s real value; and of those at further removes within financial institutions, to avoid scrutinizing what happened at lower levels of the financial food chain. The book is Deer Hunting with Jesus: Dispatches from America’s Class War. An under-appreciated work that took me two years to find.
March 12th, 2009 at 2:17 pm
Point, you’re concerns are interesting, but can be addressed fairly simply. Social science predictions leave tremendous room for variation and indeterminacy at the individual level, where free will operates. Instead, the generalized predictions social scientists make apply to aggregations, populations, etc. When we look at individuals, we treat them as hypothetical, ideal-typical, or simply illustrative. But good social science never presumes that these stylized individuals account for the full range of human behavior
Do you feel as attached to the idea of collective free will?
In the end I cannot say that social science entails any greater constraint on individual human behavior (and thereby free will) than biology.
March 12th, 2009 at 3:05 pm
“Do you feel as attached to the idea of collective free will?”
It does seem that if each individual that makes up a collective has free will, that said group could make decisions collectively — so, yes, in that one follows from the other.
(I may have to get back to you, DTM)
March 12th, 2009 at 3:35 pm
DTM (sorry about the delay; phone call, life, etc.)
Interesting point. What, though, would you say constitutes a “reliable” method? Can the method of study be reliable if the object of study is not (that is to say, not consistently predictable)?
March 12th, 2009 at 5:12 pm
Point said “It does seem that if each individual that makes up a collective has free will, that said group could make decisions collectively — so, yes, in that one follows from the other.”
Imagine the following scenario:
Tom, Dick, and Harry are going to dinner. They have to choose between Burgers, Chinese, or Pizza. Each individual has a stable, transitive preference ordering (if Tom prefers A to B and B to C, then he prefers A to C). For each individual, their ability to exercise choice according to his preferences constitutes free will.
Aggregate preferences, assign 3 points to first preference, 2 points to second preference, 1 point to third preference.
Tom prefers Burgers (3pts) over Pizza (2) over Chinese (1)
Dick prefers Pizza (3) over Chinese (2) over Burgers (1)
Harry prefers Chinese (3) over Burgers (2) over Pizza (1)
Chinese wins 6pts
Pizza wins 6pts
Burgers wins 6pts
Thus, the group has no stable, winning preference. This simple thought experiment demonstrates that even when all individuals have free will, the collective might not. Social choice theorists should recognize this as an example of (Kenneth) Arrow’s Impossibility Theorem. There are many examples in economics of the problem of aggregating individual preferences up to the collective level. See Tom Schelling’s “Micromotives and Macrobehavior” for an in-depth treatment of the general problem of individual vs. collective preference satisfaction (ie, free will).
March 12th, 2009 at 7:52 pm
pickabone
I take your point completely — any attempt to try to understand a group’s choices has to take the possibility of cancelation and second choices into account. Benjamin Constant was also known for this insight.
Nonetheless, what I hold is that groups “could make decisions collectively” — that these decisions stem from the free will of their parts (the members of the group), and so are based on free will. And thus cannot be understood by anything resembling a scientific method.
March 12th, 2009 at 7:53 pm
Oh, and thanks for the Tom Schelling reference…
March 12th, 2009 at 9:37 pm
If journalists weren’t innumerate, they would have gone to business school rather than journalism school.
March 12th, 2009 at 9:38 pm
Pickabone: Great example, but I’ve seen the conundrum explained this way:
If they vote on burgers vs. pizza (ignoring chinese), Tom & Harry favor burgers, so burgers wins.
If they vote on burgers vs. chinese, Dick & Harry favor Chinese, so Chinese wins.
If they vote on chinese vs. pizza, Tom & Dick favor pizza, so pizza wins.
Thus, the group prefers burgers > pizza > chinese > burgers (????)
The group does not have a single preference relation that obeys the transitive property.
March 12th, 2009 at 10:28 pm
“Keep in mind there are some natural systems which don’t have “free will”, but which are so complex and sensitive (aka “chaotic”) that they are not perfectly predictable (e.g., the weather).”
An interesting point, though I’m wondering if the weather is the best example of a chaotic system — because, to my understanding, the error in weather prediction is pretty much tied to lack of data (why longer term climate predictions are, as a rule, more accurate than day to day forecasts — more data can be gathered over a longer period of time).
And I apologize if completely misunderstand you, but is this what constitutes a phenomena being “complex”?
If so, I would have to differ on the point of human predictability — to my knowledge, unlike the weather, we do not become more predictable with more data.
March 12th, 2009 at 10:41 pm
Human behavior occurs in a context. Perhaps the pizza place is closer to the movie theater so they go there despite an inherent preference on the part of some (or all) for other cuisines. It is very hard to figure out the context for each individual and each situation, so prediction only applies in artificially simplified conditions, and hence is unsatisfactory in real life.
March 12th, 2009 at 10:45 pm
nbt: yes, that’s another way to express the problem, that individual transitivity yields collective intransitivity.
On a side note, one solution to the problem is agenda control. If Tom is in charge of the voting rules, he can force an election between Pizza and Chinese first:
Tom prefers Pizza over Chinese
Dick prefers Pizza over Chinese
Harry prefers Chinese over Pizza
Pizza wins first round. Next, the winner of round one goes up against Burgers:
Tom prefers Burgers over Pizza
Dick prefers Pizza over Burgers
Harry prefers Burgers over Pizza
Burger wins second round, Tom gets his preferred meal and Dick gets the shaft. By controlling the voting sequence, Tom wins. Thus, the “free will” of the collective is really the free will of the person who makes the voting rules.
March 12th, 2009 at 10:53 pm
harold, two words for you: ceteris paribus. trans: “all else equal”
In other words, controlling for relevant contextual variables, such and such finding is expected.
And on a side note: the majority of movie theaters are in commericial centers. Most of these locations have roughly equally convenient burger joints, pizza parlors, and chinese restaurants (there’s a reason I didn’t choose Ethiopian food as one of the options, even though it’s delicious). Furthermore, many movie-goers have access to cars. Therefore, the difference among travel times between the theater and the restaurant options is negligible. Thus, in this particular case, the ceteris paribus assumption is credible.
March 12th, 2009 at 11:00 pm
Just sent Matt a notice on this thread — we may have his thoughts on this tomorrow.
March 12th, 2009 at 11:03 pm
point said, “Nonetheless, what I hold is that groups “could make decisions collectively” — that these decisions stem from the free will of their parts (the members of the group), and so are based on free will. And thus cannot be understood by anything resembling a scientific method.”
I’m not sure that being “based on free will” is the same having free will, which is the condition you posit as disabling scientific inquiry. One of the main take-away points from Arrow’s Theorem is the collective decision making depends greatly on the decision rules for aggregating individual preferences (as illustrated in my response to nbt above). These rules are not fore-ordained, and themselves are the subject of much social-scientific inquiry.
One of the great things about Schelling’s book is that it shows how certain collective decision-making arrangements can yield outcomes that cut against the preferences of all of the individuals involved. Basically, some aggregating rules sacrifice the free will of some people, but other sets of rules sacrifice the free will of nearly all the people. You could make a case for the survival of collective free will in the former, but it would be hard to make the case for the latter.
March 13th, 2009 at 9:57 am
This a good point, pickabone, building on the former — and, to the degree that certain group rules reliably satisfy preferences of its members more than others on a consistent basis, they can be studied and understood by a sort of scientific method — I have to grant you that.
Still — and this is more nitpicking — at the end of the day, the groups themselves choose which rules to use — in a sense, they have a collective free will on whether they give themselves the capacity to create a collective free will.
March 24th, 2009 at 6:03 am
Great post appreciated by an Australia-based Yankees fan. G’day!
April 18th, 2009 at 6:24 pm
I like your style.