Informed Voters Choose Worse

From Time:

Political scientists Richard Lau at Rutgers and David Redlawsk at the University of Iowa have developed four models of how people actually pick candidates.  No partisan or demographic group is predisposed to a particular model, and a voter might use different strategies for different contests. …  More than 70% of the time, voters end up checking the box for the candidate who shares their views. …
Passive Voter:  You don’t look for facts about the candidates, other than their party affiliation. …
Frugal Voter:  You learn the candidates’ stands only on topics you really care about, ignoring all else. …
Intuitive Voter: You seek only enough information to reach a decision. … the process appears to be almost unconscious. …
Rational Voter: You actively seek as much information as possible about all candidates, consider the positives and negatives and evaluate them against your personal interests.

  • Because you learn so much about both sides, this strategy is highly likely to lead to a vote across party lines.
  • This strategy is also the most likely to result in a incorrect choice – picking a candidate who does not reflect your views. Researchers think that many people can’t process all they learn and simply become confused.

Source: How Voters Decide: Information Processing During Election Campaigns, by Richard R. Lau and David P. Redlawsk 2007.

Got that? Voters who try to learn more than just a few things end up less able to pick candidates who share their views.  So either humans just aren’t capable of supporting a more informed democracy, or our most "informed" voters aren’t really trying to make a good choice.  Apparently for them, voter information isn’t about voting policy. 

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • Yvain

    The Time article doesn’t say much more than you did, and I can’t find the original, but my gut reaction is that this could be an artifact of the way they measured it.

    If “how well the candidate shares my views” is measured by taking five or six hot button issues, like the Iraq war or abortion, and then matching whether the subject and the candidate are both pro-choice, or both pro-withdrawal, then the people most likely to vote for the candidate who shares their views are the people who learn just enough to know “X is pro-choice and so am I, I’ll go with X.”

    A voter who learns much more than this, examines X’s record in depth, and finds that he has a fundamentally different political philosophy than X does, or that X has a history of corruption and dirty politics, or that X has different positions on small niche issues that may be important to that particular person, might be tempted to vote against X for completely rational and correct reasons that a quick study wouldn’t capture.

  • http://www.riskanalys.is Alex

    I’ve been trying to explain this to people for years! However this:

    “So either humans just aren’t capable of supporting a more informed democracy, or our most “informed” voters aren’t really trying to make a good choice. Apparently for them, voter information isn’t about voting.”

    statement frames the conversation as if it is problem with the voter, rather than the system. Is it that I’m not capable of making a good choice, or is it that neither party can be right 100% of the time, so I have to pick and choose who is “most right” on most issues, (which, given the capabilities of our government to execute – ends up being who will do the least harm on most issues)?

  • Julian Morrison

    Isn’t this a perfectly normal result of trying to use “system 2” (but not actual Bayes) to perform a choice? As a result you drown out “system 1”, vacillate with a pantomime of the rules of logic, and end up picking nearly at random.

  • Ian C.

    The trouble is, pollies tell you what you want to hear. So some of the information out there is spin, which has a negative effect on your objectivity. More information is not always better. Certain information (voting records?) might increase your objectivity whereas other information (speeches?) might decrease it.

    It seems odd to say that less information can sometimes better, it goes against all my instincts.

  • http://zbooks.blogspot.com Zubon

    a incorrect choice – picking a candidate who does not reflect your views
    This needs elaboration. It sounds too much like “What’s wrong with Kansas”? Yvain is on point with less shallow considerations of what might best reflect your views.

    I also note that “voter information isn’t about voting policy.” Some people are interested in politics as a form of entertainment, engaging in behavior that looks like information-gathering but is not directed towards better decision-making. They already know enough to vote, and they are not following everything closely so that they can update their priors. They are watching a multi-month sporting event and rooting for their teams.

  • http://profile.typekey.com/aroneus/ Aron

    I imagine we accept that prediction markets function better than democratic ones for collective deciding since the neurons involved in the decision making are rewarded\punished for their performance. If we wanted to setup some kind of ‘lifetime voting record’ weighting (i.e. voter X counts for 10 times voter Y), how would we structure that system?

    We need a friendly election process.

  • Stuart Armstrong

    Passive Voter: You don’t look for facts about the candidates, other than their party affiliation. …

    I waste a lot of time looking at politics, but I still think that party affiliation is the most reliable piece of information about a candidate. The exact same candidate, in a different party, would not govern or vote the same way. And party affiliation is a signal nearly impossible to fake.

  • Stuart Armstrong

    So either humans just aren’t capable of supporting a more informed democracy, or our most “informed” voters aren’t really trying to make a good choice.

    I think they are suffering from the dellusion that since they have accumulated so much extra information, they have to make a decision that reflects all of it – not just he few important facts. And that since the info is complicated and contradictory, their decision should be complicated and contradictory.

  • http://yudkowsky.net/ Eliezer Yudkowsky

    There are known experiments where humans seem to do worse by gathering more data. But if this is not an experiment, but a computer model, I really want to know how their “Rational Voter” works, and what super-rational gold standard that voter is being compared to by which they’re said to fail. Sounds very suspicious. If it’s not a computer model, I still want to know what gold standard they’re comparing information-hungry voters to.

  • Michael Bishop

    I agree it would really help to have more information about the study. I started looking for it but have to stop now. If you’re interested, perhaps these links are decent places to start:

    http://scholar.google.com/scholar?q=Richard+Lau&hl=en&lr=&btnG=Search

    http://www.uiowa.edu/~c030111/personal/personal.html

  • Aaron

    Alex makes a good point. It’s something I’ve been thinking about more after reading “Gaming the Vote”. If it was a matter of rating each candidate as opposed to ranking them, rational voters might be able to do a better job.

  • Fletcher

    ..

    1) It is impossible for any voter to understand all the relevant policy issues in large state or national elections.

    2) It is impossible for any voter to discern candidate(s) true views on any/all relevant policy issues.

    3) It is effectively a zero probability that any voter will determine the outcome of any large election.

    4) It is illogical,therefore, to vote in large elections, as a rational cause & effect human action.

    “Irrational Voting” {..’emotional voting’} cannot be avoided in any but the smallest local elections.

    Efforts to better ‘inform’ oneself on issues/candidates are futile… the Signal-to-Noise-Ratio decreases exponentially as a voter attempts to collect & sift large amounts of conflicting and often unreliable “data”. Confusion results, rather than rational enlightenment… “informed voters” do worse.

    Large-scale “informed-democracies” are logically impossible… they are oligarchies in the real political world.

  • http://uncommon-priors.com Paul Gowder

    How did the frugal voter do? The frugal voter sounds more rational than the “rational voter” …

  • Anonymous

    Efforts to better ‘inform’ oneself on issues/candidates are futile…

    Worse than futile, they waste your precious time better spent elsewhere.

  • Grant

    Putting aside critiques of the study’s methodology, I think some of this is just what the voter’s demand.

    Because voter’s don’t demand rational, contradiction-free information on each candidate, the media doesn’t supply that sort of information. It supplies partisan (e.g., Fox News) and sensationalist data which makes it nigh-impossible for any real political truth-seeker to figure out what the hell is going on. I think if the majority of voters were truth-seekers (not going to happen, I know), the information would be much, much easier to process.

    Put another way, imagine if the learning materials which covered other complicated topics (such as solving differential equations or computer programming) were all made for irrational people instead of largely truth-seekers. How difficult would it be to really learn from that material?

  • http://shagbark.livejournal.com Phil Goetz

    I agree with Yvain – Being a “rational voter” implies, by definition, that you vote using information not easily available to someone conducting a study.

    Aron makes a good point; I also am interested in applying neural network learning methods to voting.

  • Peter

    So either humans just aren’t capable of supporting a more informed democracy. . .

    Democracy is not synonymous with electioneering, nor elections in general. It doesn’t even mean that a country has to hold elections.

    Maybe the correct interpretation of this data is that, even with an informed populace, the concept of representative democracy is inherently flawed.

  • Grant

    Maybe the correct interpretation of this data is that, even with an informed populace, the concept of representative democracy is inherently flawed.

    I always find it interesting that sociologists who read and believed Hayek’s notion of dispersed, tacit knowledge could turn around and support democratic outcomes. It seems obvious to me that democracy and the division of knowledge cannot occupy the same space.

    I’m not seeing this as a flaw in humans necessarily. Specialized intelligences of any sort should encounter similar problems in democracies, shouldn’t they?

  • http://www.spencertipping.com Spencer Tipping

    To some extent this phenomenon makes sense. If you know one fact about a candidate, for example that you agree about gun control, then they’re positive. However, if you then learn that you disagree about something far less important, for instance, ten years ago they voted for legislation to change taxes in a way that you didn’t like, they may be demoted to “neutral” status again.

    People inherently have a difficult time managing levels of significance. The more you learn about someone, the more you have to do this. If you learn only the major positions that a candidate takes, you have a more accurate representation of whether you’ll agree on important issues.

    Malcolm Gladwell’s book Blink discusses this concept in many different forms; I highly recommend it.

  • Tom Breton (Tehom)

    What jumps out at me is that nobody would actually do what the category “rational voter” says. It says:

    You actively seek as much information as possible about all candidates

    Nobody would do that. There’s just too much information out there. Nobody would, say, read political news on Google News 24×7 just so they could vote well.

    Sounds good, though, doesn’t it? Why, you haven’t left out any information you possibly could get. None at all. I think this category appeals to a certain sort of posturing that wants to be above reproach.

    I have to suspect that the category “rational voter” was full of people who placed themselves in it for the wrong reason.

  • Peter

    I always find it interesting that sociologists who read and believed Hayek’s notion of dispersed, tacit knowledge could turn around and support democratic outcomes.

    Maybe the ones who actually believed Hayek were all intellectually dishonest, insane, or just not very analytical?

    Specialized intelligences of any sort should encounter similar problems in democracies, shouldn’t they?

    I suspect that you’re still talking about voting and electing leaders, and equating this with democracy (e.g. a system in which a lot of people vote and members of the electorate have specialized intelligence…). Correct me if I have you wrong.

    Only a handful of mature democracies exist on the planet – Switzerland, and the Nordic countries which are de facto democracies. These countries are also home to some of the most intelligent and educated populaces on the planet.

    Maybe large amounts of information and knowledge is a problem in the U.S. (if this study says so), but one of the most annoying things one can do is extrapolate good data from the American system and try to apply it as a general rule. America’s government and voting schema is incompatible with a knowledgeable electorate – who knew?!

  • http://google.com/ Daniel Reeves

    My hypothesis is that “informed voters” weigh all issues more equally than they should. For example, let’s assume the following are the only three issues, war being the biggest one:

    – Voter A is for the war, but against abortion rights and against gay marriage
    – Candidate X is for the war, for abortion rights, and for gay marraige.
    – Candidate Y is against the war, against abortion rights, and against gay marriage.

    War here is obviously the biggest issue. If Voter A is uninformed and only knows the candidate’s positions on the war, then he’ll vote for Candidate X. But if he decides to research the positions more… who knows what will happen! If Voter A irrationally weighs all positions equally, then he’ll vote for Candidate Y.

    The thing is, not all issues are equal. Has Bush ever even voted for a bill on gay marriage? No. Abortion rights? Well, he’s only denied some stem cell research funding, but again, that’s not a big issue when compared to the war. Staying uninformed means that you’ll only get to hear the candidate’s views on the big issues, and all things considered, it’s what your vote should boil down to.

  • Grant

    Peter,

    Yes, I meant a representative democracy, though I think the point applies to a direct democracy as well. Even if voters weren’t biased and only voted on issues they felt educated on, they’d still have to be incentivized and able to accurately identify which issues they were qualified to vote on (which I think may be a problem this study highlights: How does a voter know when he or she is informed?). Absent policy prediction markets I’m not seeing how this could work well.


    America’s government and voting schema is incompatible with a knowledgeable electorate – who knew?!

    I agree its government supports irrational voting, but we couldn’t we also say its electorate cannot elect a rational government with a decent voting schema? These factors seem self-reinforcing.

    Are there any studies showing less voter irrationality in countries with direct democracies, as you seem to be suggesting?

  • http://www.riskanalys.is Alex

    “War here is obviously the biggest issue.”

    Says you!

    (sorry, couldn’t resist)

  • http://goodmorningeconomics.wordpress.com jsalvati

    I find this hard to believe, and I’ll do more investigation. I find this hard to believe because, Bryan Caplan is convincing in telling us that voters normally make bad decisions, and more education leads to more reasonable choices.

  • Lars

    “Efforts to better ‘inform’ oneself on issues/candidates are futile…

    Worse than futile, they waste your precious time better spent elsewhere.”

    I think people are missing the most important news filters of all: You. You can filter the noise and get a good grasp on policies.

    Most rhetoric, like speeches, can safely be ignored or taken with a huge grain of salt. Anyone trying to get elected has to say things to appeal to certain demographics of the population without offending other demographics too much, so as to appeal to the widest variety of people as possible.

    For example, in most press, Barack Obama talks about ‘main street and wall street’. You know, he’s for ‘main street’ and against ‘wall street’. What does that mean? Nothing. It doesn’t mean anything. It’s just something he says in speeches to hopefully appeal to a broad group of people who don’t like ‘wall street’, whatever that means. Big business? Both mainstream candidates got most of their campaign money from big companies, so it’s definitely not that. There’s plenty of examples of rhetoric like that from McCain or any other politician, of course.

    Personal stories about the candidates and profiles of their kids and how much they weight and how tall they are and where they shop and all this trivial information is easy to get drawn into, but it doesn’t mean anything. All that news is pushed at the general population because we humans are drawn to people, not ideas (there was a recent post here about this).

    The point is, ignore the rhetoric. Actions speak louder than words. If you want to see what a person will do in the future, look at what they did in the past. Look at where the money is coming from, look at what that group or political party has done in the past, and look at history to get an idea of what’s going on. That’s a lot of work, and every source of information has their own agenda, so it’s up to you to weight things as you see fit.

  • http://profile.typekey.com/michaeljameswebster/ michael webster

    Here is another spin on the topic of rational voting:

    “The founders of the United States didn’t have the advantages of fMRI brain imaging and had no concept of the amygdala, but they were hesitant about political parties and political campaigning nonetheless. Fearful that a “torrent of angry and malignant passions will be let loose,” Alexander Hamilton railed against political parties in the first Federalist Paper, saying the parties would try to “increase the number of their converts by the loudness of their declamations and the bitterness of their invectives.”

    It turns out there was some reason to be concerned about the relative influence of information versus emotion when it comes to political judgments and affiliations. Though it is impossible to know for sure whether people actually vote along party lines, many psychological studies have shown that political affiliation plays a large role not just in the voting booth but also when people must decide how they feel about political issues.”

    From: http://www.psychologicalscience.org/observer/getArticle.cfm?id=2393

    Geez, I don’t recall any of these empirical insights being expressed as axioms that a social choice function has to satisfy.