A politician can influence voters just by stating an opinion, even an opposing opinion

There is a new research paper by David E. Broockman and Daniel M. Butler (pdf), the abstract is this:

Politicians have been depicted as, alternatively, strongly constrained by public opinion, able to shape public opinion if they persuasively appeal to citizens’ values, or relatively unconstrained by public opinion and able to shape it merely by announcing their positions. We conduct unique field experiments in cooperation with legislators to explore how constituents react when legislators take positions they oppose. For the experiments, state legislators sent their constituents official communications with randomly assigned content. In some letters, the representatives took positions on salient issues these constituents opposed, sometimes supported by extensive arguments but sometimes minimally justified. Results from an ostensibly unrelated telephone survey show that citizens often adopted their representatives’ issue positions even when representatives offered little justification. Moreover, citizens did not evaluate their representatives more negatively when representatives took positions citizens opposed. These findings suggest politicians can enjoy broad latitude to shape public opinion.

I suppose Alex Salmond is one current leader who understands this, Putin is another.

For the pointer I thank the excellent Samir Varma, who also cites coverage from Wonkblog.

Comments

File under: some people are actually skilled politicians

Also, there is no great stagnation

Also, bravo for that second-to-last sentence. Outstanding.

If you go all the way to page 42 of the paper, you can see a graph of the actual results. It shows that the size of the effect is quite small and just barely significant. None of the individual issues were significant in themselves, but the ones with the largest positive effects were relatively uncontroversial or minor policies where people may have no formed opinion.

"Broad latitude to shape public opinion" seems like a real reach.

I think it's cheating to look at the actual results of papers commented upon here. Especially when you discover that effects are generally small, ieven when they are significant. Subjecting papers to that kind of impossibly high standard undermines social science research.

+1

Dan111 and gang,

The major finding of the paper is that there is not a NEGATIVE effect on voters' opinion of their elected officials when they find out that their official supports a policy that the voters oppose. So, a small positive effect makes their finding that much stronger, especially since voters know very little about their elected state legislators in the first place.

Their last line could be seen as an overstatement, but they don't say "we find." Rather, they say that their findings "suggest" and after reading the paper, that seems quite accurate.

Figure A2 on page 42 shows a negative effect on legislator favorability for eight issues surveyed, and a positive effect on nine.

In light of that, this statement in the abstract is both true on average and also deeply, purposefully misleading:

Moreover, citizens did not evaluate their representatives more negatively when representatives took positions citizens opposed.

In fact they did on nearly half of issues surveyed!

Figure A1 on issues, rather than favorability towards the politician, seems like a better bet to make a strong statement on. But that ain't the one the authors chose for headline results in the abstract.

@a Michael, that is a misreading of the paper and is clearly contradicted by the excerpt quoted above. The paper is mainly framed as "Do legislators influence voters' opinion?", and their main finding is a claim that they have found evidence of such influence. I believe the main finding is overstated.

Yes, there is also a finding that taking an opposing position has no negative effect on voters' position. However, that is a different analysis from the main finding.

@Dan111, where in the abstract does it say that the paper is framed as "Do legislators influence voters’ opinion?"

The abstract does say, however "We conduct unique field experiments in cooperation with legislators to explore how constituents react when legislators take positions they oppose." And that's exactly what the paper does.

They only time they mention legislators' influencing voters' opinion is in the first sentence, which is presenting a claim made by scholars and political observers, and in the last sentence, which does relate to their own findings and is hedged with the words "suggests."

Are we reading different abstracts?

The studies with the strongest negative effects have very small sample sizes (less than 50). I wouldn't characterize the abstract as misleading; only open to interpretation.

In Figure A2, the mean and median sample size is 64 and 54 for team positive. It's 51 and 52 for team negative.

If you want to criticize the study on sample size, then throw out the whole thing. Or, if you want to take the positive effects seriously, then take the negative ones seriously too.

If I had a study where I gave a drug to seventeen groups of people, and eight of those groups became ill, do you think I ought to say in my abstract, "Moreover, subjects did not suffer negative effects from the drug"?

Or do you think that would be..... misleading?

you are too kind. This stuff is pure speculation with no factual basis at all, but solemnly presented here as a serious "new research paper". Professional academic seem to thrive on this dominant type of research nonsense.
(Publish something/anything or Perish)

Why do you call it pure speculation? They did an experiment and reported results. Seems like about as good a test as one could think of.

Ever heard of statistical significance?

Speculative, perhaps, but clearly not "pure speculation".

Perhaps you have something constructive to add, such as some refinement to the methodology which would lead to superior statistical virtuosity, or like the point you responded to, which informed us of that magnitude and significance of the results.

Perhaps "some latitude" would have been more consistent with the study results.

"Publish something/anything"...

You call getting actual elected officials to successfully randomize their political communications with their constituents, which lies at the heart of the relationship between voters and their representatives, and then conducting large surveys of the voters before and after the treatment "anything"? Not only is this a very important question and a top-notch research design, but it's also a very costly project in terms of money, time, and risk (most field experiments fail in the implementation stage, many others result in non-interesting findings). Please tell me, what does good empirical research look like, then?

The study design really was excellent. Full points. And I'd like to know how the heck they convinced politicians to participate!

But the best study design in the world doesn't justify a misleading abstract.

What is misleading about it? The part when they say that some people argue that politicians' can lead public opinion and then state that they test out how voters respond to learning that their politician differs with them on a policy issue? Or is it the part when they say that their findings "suggest" that politicians have broad latitude? I thought their findings suggested as much.

Read my comment at 8:04am

@mosherT, have you read the paper? I have at least skimmed through it. It appears to be a well-designed experiment on the topic. The problem is just their exaggerated interpretation of the findings.

> @dan1111:

No, it is not "a well-designed experiment on the topic" ... but yes, it is an extremely "exaggerated interpretation of the findings". They didn't statistically measure anything whatsoever.

Flaws in the design are huge and stunning. Not even an attempt at actual random sampling, though the authors sprinkle the word 'random' in their report to fool the naive. They cannot possibly generalize their tiny, sloppy experimental results to the much broader statistical populations of "citizens" and "politicians" ... but they arrogantly do so anyway.

Their methodology and terminology is pathetic: "We conducted our experiments in collaboration with eight Democratic state legislators
from a Midwestern state. We conducted the first experiment in the summer of 2013 with one of these state legislators and the second experiment in collaboration with the other seven state legislators in the spring of 2014 .... We surveyed 1,210 voters in Legislator A’s district in May 2013 and asked them about their positions on these issues in order to determine who would be included in the experiment’s sampling frame." (does that sound like random sampling to you ?? -- heavily biased selection of 8 Dem Pols in a single state, with a cherry-picked group of their district "voters" for a highly contrived and error loaded survey procedure)

This "research report" (76 pages) is complete nonsense from both statistical and common sense viewpoints.

It doesn't appear that you understand randomization.

Random sampling of the entire population is not necessary and almost never feasible in experimental research. Further, it is not necessary that the study population have the same distribution as the population about which one is making inferences, only that all groups one wants to make inferences about are adequately represented in the population. For example, a study of 600 women and 200 men would not be biased just because the distribution isn't 50% men and 50% women; as long as 200 men is an adequate sample size, you can safely apply your results to both men and women.

Selecting only Democratic voters might be considered a problem, because then someone could posit that Republican voters behave differently; therefore, the results could only be applied to Democrats. However, selecting only Democratic constituencies is not a problem, because these areas still have large numbers of Republican voters (in some districts the vote for Obama was as low as 50%). Now, there are always arguments to be made about representativeness of the population--these districts were all in one state, and maybe voters in this state are different from others, etc. However, this is no different than any other experimental research. Any experiment uses a particular population, and that population often has to be selected based on factors of cost, accessibility, etc.

The important randomization was randomizing which people received which statements of the legislators' positions--which they did.

> dan1111:

you're getting lost in irrelevant trivia and making nonsensical assertions about randomization.

Get back to basics to organize your thinking in statistical probability sampling. From this report:

>> precisely state the hypothesis/theory that is (supposed) to be tested by this research project

>> precisely state the statistical "population" to be sampled in order to disprove/prove this hypothesis

>> state the general requirements for a statistically representative sample of that specific population being studied

Note that the authors of this research study didn't bother with such fundamental statistical detail, thus earning criticism.
One cannot scientifically attribute conclusions to a large general population... from a small sample of that population if rigorous statistical principles are not followed. Confusion about what is actually being sampled/tested is extremely common, especially in social science research.

I remember a Robin Hanson post where, in the comments, he decided to give up reasoning from abstracts. I think he used an abstract, made a number of insightful comments about it, and then a reader pointed out that the paper's results didn't support the abstract at all. Given how many times that had happened before, Robin declared all abstracts suspect until proven otherwise.

I think academics have cottoned on that most papers never get read, and so make outlandishly interesting claims in their abstracts. It paid off in this case - with no less a prominent blog as MR re-broadcasting it! - until dan1111 all did us a fantastic service by reading to danged page 42.

Well done Dan1111.

What's worse, believing the abstract or some commenters unfounded statements about the abstract?

Thankfully, those aren't the choices we face today. dan1111's page 42 pointer was founded.

So, yet more evidence that the "opinions" of the masses are not the primum movens of our "democracy" but instead lie downstream of the loci of true political power (as always).

"A politician can influence voters just by stating an opinion" Isnt this result a bit too... obvious?

In other news.... the median voter is found to be a complete sheep.

"Your representative owes you, not his industry only, but his judgment; and he betrays, instead of serving you, if he sacrifices it to your opinion."

http://press-pubs.uchicago.edu/founders/documents/v1ch13s7.html

Tribal loyalty. Humans haven't changed much.

Putin a politician? I guess some times and in some ways, maybe. It's kind of like grouping Al Capone in with "liquor distributors".

Any opportunity for a dig at Scottish Nationalists, eh Tyler. Or are you just trying to influence me by stating an opinion?

Hello! I know this is kinda off topic but I was wondering which blog platform are you using for this site?
I'm getting fed up of Wordpress because I've had problems with
hackers and I'm looking at alternatives for another platform.
I would be fantastic if you could point me
in the direction of a good platform.

Kim Kardashian certainly isn't the first celebrity to get some work done, but the 'Keeping Up
with the Kardashians' episode did prompt a plastic surgeon, Dr.
She left the doctors at the second hospital mystified as well.

Vanity is when you have something taken care of because you want everyone to notice and look at you.

It's hard to find well-informed people about this subject, however, you seem like you know what you're
talking about! Thanks

Go through Slendertone review and you're going to get to know more magical facts concerning this improbable
product. You are to plan your life ahead and move on with that.
The taste alone may be enough to keep you from wanting to overeat.

Click onto the Parlour X website ASAP & discover
this fabulous exciting new shopping experience'.

Of course, you can dress like a celebrity if you want to,
but you need to remember that celebrity fashions regularly change; therefore, there is no need for you to go broke trying to afford the hottest,trends that may
only last for a few weeks. If you are under doctor's orders
to limit physical activity it is unwise to try to play basketball unles advised otherwise.

Se volete a crescere conoscenza appena sito ed essere aggiornato con le ultime calde informazioni postato qui .

Now when it comes to the actually tweezing part, you
want to make sure that you're using a good
pair of tweezers. The platinum blonde is
a much sought after persona, and the look, an elegant
and beautiful one if carried off properly, is a
much desired and much employed part of dressing up.
* Ginkgo Biloba is an herb which has properties to help improve blood circulation to the scalp.

The revealing and additionally strapless mermaid wedding dress outline is the most famous one.
Duty up with the current pattern trends is not as hard as it may
seem all it takes is a few tricks and a bit of investigate.
There is a separate section for skin and body and of course,
the photo galleries.

Surveys have shown that some storage and disposal firms lack the equipment to destroy the
documents themselves, and often resort to hiring subcontractors to destroy the data.

The packers are skilled and are capable enough to
pack different types of goods. The Street Fiction ebooks available at this store are one of the best
in their genre. You can find companies offering packing services and
this means that you have less to worry about when moving.
Another point to be considered here is that the local businesses know
the target customers better than multi national companies, as they have
their roots here in this county of Florida and thereby know the people and the tourists
of this place better. Functional and striking, track lighting fixtures can be found in different shapes and sizes according to your choices and needs,
from ultra modern to that which resembles turn-of-the century, any style to fit any dcor.
Middle East region is hugely popular among expatriates as a career
destination as people from all nationalities and religions come together to earn attractive salaries in return for
their expert services. Having a vast array of cruisers and sports boats, known as Stingray boats available
in different models, every model is new and distinct from the other one.
3rd party logistics providers don't just stock and
handle merchandise, they provide crucial support for shippers and retailers who are looking into overseas markets.
Professional packers and movers of Delhi are
completely safe and trouble option of relocation.

Ζahle wass burned in 1777 and 1791, and it was burned and plundered
iin 1860. His precise, forceful fills propel the band forwar with սndeniable power.
Ouur gսіde on site iss a short, elderdly Turƙisɦ gentleman fluent in English, a retired history
teacҺеr.

Comfort comes in the form of the beautifully upholstered
reclinable deck seats and bow seats, and a drive. (AND "FLAVOUR, FEEDING TRIGGERS AND CHEMORECEPTION SECRETS") SEE:.

Bait casting are multiplying reels, with the spool mounted above the rod.

Have you ever thought about writing an e-book or guest authoring on other sites?
I have a blog based upon on the same ideas you discuss and
would love to have you share some stories/information. I know my
audience would appreciate your work. If you are even remotely interested, feel free to shoot me an e mail.

Hello there I am so glad I found your website, I really found you by mistake,
while I was browsing on Digg for something else, Anyways I am
here now and would just like to say cheers for a incredible post
and a all round entertaining blog (I also love the theme/design), I don’t
have time to read it all at the moment but I have bookmarked it and also added in your RSS feeds, so
when I have time I will be back to read a great deal more, Please do
keep up the excellent jo.

bed rails for seniors walmart

Quality posts is the important to interest the viewers to pay a visit
the website, that's what this web site is providing.

Comments for this post are closed