Skip to content
 

Where does the discussion go?

Jorge Cimentada writes:

In this article, Yascha Mounk is saying that political scientists have failed to predict unexpected political changes such as the Trump nomination and the sudden growth of populism in Europe, because, he argues, of the way we’re testing hypotheses.

By that he means the quantitative aspect behind science discovery. He goes on to talk about the historical change from qualitative to modern quantitative analysis which hinders the capacity of scholars to study ‘less common’ or unfrequent situations, such as the ones outlined above.

I [Cimentada] am pretty sure there’s some truth behind that, but still, I think that the capacity to predict is not entirely based on the frequency of things. Another thing he fails to distinguish is that specific questions require specific designs. Depending on your research question, you might need to use qualitative over quantitative approaches.

If you have some time, I’d like to hear your stand on this. I thought this might be something which could fit in one of your blog entries which is why I contacted you.

I leave you with one paragraph that summarizes his main point pretty well:

It is easier to amass high-quality data, and therefore to make “rigorous” causal claims, about the economy than about culture; in part for that reason, the social sciences now favor economic over cultural modes of explanation. Similarly, it is easier to amass high-quality data, and to test causal hypotheses, about frequent events that are easy to count and categorize, like votes in Congress, than about rare and intractable events, like political revolutions; in part for that reason, the social sciences now tend to focus more on the business-as-usual of the recent past than on the great turning points of more distant times.

My reply: these are good questions that are worth considering. It’s kinda funny that they appeared in an opinion article in the Chronicle of Higher Education rather than in a political science journal, but I guess that journals are not so important for communication anymore. Nowadays journals are all about academic promotion and tenure. When people want to have scholarly discussions, they turn to newspapers, blogs, etc.

23 Comments

  1. Matt Skaggs says:

    “…it is easier to amass high-quality data, and to test causal hypotheses, about frequent events that are easy to count and categorize, like votes in Congress, than about rare and intractable events…”

    This sentence is problematical. The statement that “it is easier to amass high-quality data about frequent events than rare events” is trivially true. The statement that “it is easier to test causal hypotheses about frequent events than about rare events” is, in my experience anyway, false.

  2. Terry says:

    In this article, Yascha Mounk is saying that political scientists have failed to predict unexpected political changes such as the Trump nomination and the sudden growth of populism in Europe, because, he argues, of the way we’re testing hypotheses.

    The Trump nomination is a good focal point for improving political predictions.

    The data was all there in plain sight. For many years, opinion polls showed overwhelming opposition to increases in immigration and they showed substantial support for reducing immigration. Why should it have been a surprise that a candidate would arise who took advantage of this?

    Trump was the only candidate who full-throatedly addressed the immigration issue. For a while, early in the primaries, he was the *only* candidate who took the immigration issue seriously. So why was it not obvious that he had a real shot at the nomination? (Trump’s many manifest flaws, of course, meant that he, himself, was a long shot, but why was it not obvious a candidate might arise who took immigration seriously?) As the primaries went on, and the other Republican candidates refused to steal the immigration issue from Trump, why was it not obvious that Trump was a serious contender? (Ted Cruz only belatedly tried to steal the issue, and Rubio’s attempts to steal the issue were comical.)

    Perhaps political scientists could improve their predictive abilities by not ignoring the sympathies of large parts of the electorate on selected issues. Or, to put it another way, perhaps political scientist should look outside the Overton window occasionally, because the Overton window is constantly shifting.

    Ditto for the “sudden growth of populism in Europe”.

    • Matt says:

      Well, how do you explain the polling results? It’s nice to put an explanation like this behind Trump’s nomination, but now you need to explain to me why political scientists and commentators should have ignored the polls, and poll aggregators, that were all indicating that Clinton was going to win. As has been said before on the blog, it’s not like there was a seismic shift in sentiment required to generate this year’s election results; small shifts in key electorates made a big difference. With the information available, Hilary was the *correct* choice for those making predictions. Anybody who predicted Trump was going to win needs to explain their likely voter model, or something, that allowed them to interpret the polling numbers in a way that made Trump the favorite. Otherwise I’ll assume you just got lucky.

      • Terry says:

        As to the polls, I agree with Andrew and 538. The polls were only a few percentage points off. Even if the polls hadn’t been a few points off, it would have still been a close election. But, a lot of people fooled themselves into thinking otherwise.

        But, polling is beside the point here. Mr. Mounk claims Trump’s rise reflects a huge political change. In his thinking, Trump should have never gotten close to the nomination and should have been defeated in a landslide, so a few percent in the polls is not the issue here.

        All this is consistent with my point that Mr. Mounk was surprised by Trump because they ignored the salience of the immigration issue to many voters. Once you recognize this lacuna, you see that the Trump phenomenon is not very surprising.

        This suggests that poli sci could learn something here, and think about considering why the immigration issue was suppressed, and think about ways to identify other suppressed issues that might be seized by populists, and think about when suppressed issues break through and when they remain suppressed.

        • Andrew says:

          Terry:

          I don’t think the immigration issue was “suppressed.” It was in the news for a long time. There are a lot of issues and just one election; not every issue will become a focal point.

          • Steve Sailer says:

            Dr. Jason Richwine got fired from the conservative Heritage Foundation when it was discovered that his Harvard doctoral dissertation documented the low average IQ of the children of immigrants from Latin America.

            That seems like a pretty flagrant example of the kind of suppression of the social science realities of immigration, no?

          • Terry says:

            Valid point. I mis-spoke.

            I meant that support for immigration restriction was suppressed. Thefore, the issue itself was not suppressed, but certain viewpoints on the issue were suppressed.

            • Martha (Smith) says:

              I interpreted your original remark as saying that support for immigration restriction was not adequately addressed in the polls. Does this at least partially address the issue, or am I totally off base?

              • Terry says:

                That’s not quite right.

                I think that the polls (at least the honest polls, there are dishonest ones tailored to downplay support for immigration restriction) correctly show that there is strong opposition to more immigration and quite a lot of support for reducing immigration from current levels.

                What I think is being suppressed is the political expression of this sentiment. Immigration is strongly opposed by moneyed interests such as billionaire tech titans that find native workers too expensive, and by the left, which disapproves of the behavior of native voters. Media are also strongly against honestly reporting on immigration restriction.

                Until Trump, there were almost no politicians speaking out against current immigration levels. Three attempts at immigration amnesty since 2000 were defeated, not by the opposition of politicians, but by grass roots uprisings.

                Hence, there was a big opportunity for a politician to take advantage of this voicelessness. Trump saw it and grabbed it and ran with it, all the way to the presidency.

                Opposition to affirmative action is another issue with widespread support, but no political voice. Don’t be surprised if that is the next “unexpected political change” to blindside Mr. Houck.

      • Anoneuoid says:

        Andrew had a couple posts on this:
        http://andrewgelman.com/2016/11/09/explanations-shocking-2-shift/
        http://andrewgelman.com/2016/11/09/polls-just-fine-blue-states-blew-red-states/

        I guess his position is largely that there is nothing to explain beyond why political scientists and commentators ignored/understated the uncertainty in the polls.

        Personally, I saw a lot of “politically active” people get really riled up about Trump and “spew drivel” like “F*** Trump”, etc. Combined with the simple argument that aristocracy is bad (two bushes, now two clintons, a pattern is forming here…, etc), I think this really helped Trump’s cause amongst people who don’t usually get very involved in politics (and probably won’t bother with things like polls).

        Also, in states where the outcome was pretty much pre-decided these same people wouldn’t bother to vote, so it would only show up in the swing states.

        • Andrew says:

          Anon:

          It is not true that there were big swings only in the swing states. See the graphs in our 19 things paper.

          • Anoneuoid says:

            Thanks Andrew, I haven’t looked at this data and was misinterpreting this line “small shifts in key electorates made a big difference” from Matt’s post. So it looks something like this: https://image.ibb.co/cg05yk/2016election.png

            R code:

            n = 50
            e = 8
            forecast = runif(n, .3, .75)
            actual = forecast + forecast^e
            actual = actual + runif(n, -.03, .03)

            par(mfrow = c(2, 1))
            plot(forecast, actual, panel.first = grid())
            abline(0, 1)
            abline(v = 0.5, h = 0.5)

            plot(forecast, 100*(actual – forecast), panel.first = grid())
            abline(v = 0.5, h = 0)

            All I did is add the forecasted value to the 8th power and “random” error between +/- 3%. No idea what that would mean and obviously it breaks down when the forecast gets too high… probably nothing but it roughly describes the data.

            Also, can you share your data source? It differs from the results on Wikipedia[1] (eg, WV is at 68.5% and WY is 67.4%, while in your plot they are both over 70% and WV is higher than WY).

            [1]https://en.wikipedia.org/wiki/United_States_presidential_election,_2016#Results_by_state

            • Anoneuoid says:

              Typo: “in your plot they are both over 70% and WY is higher than WV

            • Andrew says:

              Anon:

              I don’t remember where I got the numbers but it was right after the election. And, since then, some votes have been added.

              For example, for West Virginia, I had 186K votes for Clinton and 482K votes for Trump; Wikipedia has 189K for Clinton and 489K for Trump. For Wyoming, I had 55K and 174K, and Wikipedia has 56K and 174K.

              Kind of amazing how lopsided those results are. I mean, sure, I knew it, but seeing the numbers right there really brings it home.

              Anyway, I guess I should update my data and redo the graphs.

            • Andrew says:

              Also, I’m plotting Trump’s share of the 2-party vote, not his share of the total vote, in each state.

  3. i says:

    I don’t have a subscription to the Chronicle so I’ll take Cimentada as an accurate representation of Mounk’s piece. A couple of things come to mind. First, it is not like political science is all quantitative. There is still a whole lot of qualitative scholars around. But I haven’t noticed them being noticeably more successful in predicting ‘big events’. Second, political science was for a long time dominated by qualitative scholars taking approaches that seem comparable to those advocated by Mounk. Again, it doesn’t strike me that ‘political science’ was any more successful in predicting unexpected changes (in which case they would, of course, no longer be unexpected).

    The main reason this debate doesn’t take place in scholarly journals is because it is old and is covered in first-year graduate courses in any decent Ph.D. program. That is not to say that there are no further contributions to be made but as far as I can tell Mounk’s article doesn’t offer anything new to the debate

  4. Terry says:

    I don’t know much about academic political research, but this statement in Mounk’s CHE article gives me pause.

    Over the last three decades, political scientists have made impressive strides in understanding how the political world functions …. Scholars of the judiciary have shown to what degree the past ideological alignment of Supreme Court justices predicts their decisions in future cases. Scholars of Congress, meanwhile, can anticipate with great accuracy how representatives are going to vote on a pending bill by looking to such factors as their party ID, their past votes, or their biggest donors.

    Is it really true that political scientists consider these findings to be “impressive strides”? They seem pretty obvious to me. Is Mounk unintentionally slighting the subtlety and power of poli sci research?

    Second, political science was once dominated by scholars with a proudly qualitative approach…. Today, by contrast, political scientists predominantly use quantitative methods. Most of them spend the bulk of their time on constructing data sets and using statistical techniques to identify how a particular set of factors is associated with — or, preferably, causes — the outcomes in which they are interested.

    Is the Trump phenomenon really outside the reach of empiricism? Couldn’t you empirically test a theory that says that strong voter opinions neglected by mainstream politicians are sometimes used by outsider politicians to gain prominences? One could collect data on voter opinions, identify “orphan” opinions with little political support, and see if they sometimes are used by outsiders. (An “outsider” doesn’t have to be completely outside politics, but could be generalized to include all lesser politicians.) The theory could be generalized to state that less-prominent positions are more likely to be championed by less-prominent positions, and more-prominent positions by more-prominent politicians.

    … For that reason, the failure to foresee the emergence of Trump wasn’t a lacuna in the enterprise of contemporary political science that might easily be addressed. On the contrary, it was a direct result of its core commitments.

    Is Trump’s rise really so abnormal, or is this cocoon-talk? Nativism was very strong in the 1920s following the waves of prior immigration. There have been many populist politicians from the political fringes, such as Andrew Jackson, William Jennings Bryant, and Bernie Sanders. Have the times really ceased to be “normal” or is it just that Mr. Mounk’s worldview is a bit narrow?

    Extreme candidates who attack basic democratic norms do not gain mass support among members of the American (or Polish or Swedish) public. … liberal democracy is there to stay. … Liberal democracy is entering a deep crisis. … It is bitterly ironic that the methodological turn in political science has taken place at the very moment when the new set of tools it mandates is especially likely to blind us to the radical transformations going on around us. … the fate of liberal democracy hangs in the balance

    Uh oh. Mr. Mounk has been drinking the Kool Aid. He thinks a populist politician democratically elected by appropriating popular positions is a threat to the foundations of democracy.

    Mr. Mounk might profit from reading about the history of western European democracies in the post-WWII period, such as Italy, Greece, and France, and even Britain. The current times are far from abnormal by historical standards.

    • Steve Sailer says:

      What has been going on is that the Establishment dogma on immigration has been getting ever more extreme in recent years, with more and more respectable authority figures implying that the citizens of a country no longer have the right to discriminate against foreigners in keeping them out. Due to heavy suppression of debate over immigration, this growing open borders fundamentalism wasn’t all that noticeable until the German Chancellor in August 2015 suddenly decided, upon a non-democratic whim, to let in a million people in violation of EU rules.

      This shocking 2015 event, much praised at the time by Establishment mouthpieces, led to 2016’s Brexit and Trump victories.

  5. Terry says:

    Mr. Mounk does not appear to be a serious political scientist. Last month, he wrote in the NYT that:

    America is on its way to a full-blown constitutional crisis.

    Over just a few days last week, President Trump and his allies stepped up attacks on Robert Mueller, the special counsel investigating the campaign’s connections to Russia. They tried to push Attorney General Jeff Sessions out of office. They thought out loud about whether the president can pardon himself.

    This all points to the same conclusion: Mr. Trump is willing to deal a major blow to the rule of law — and the American Republic — in order to end an independent investigation into his Russia ties.
    https://www.nytimes.com/2017/08/01/opinion/trump-democracy-institutions-destroyed.html?mcubz=1

    Mr. Mounk does not seem to understand that Mr. Trump is the president and that Mr. Mueller and Mr. Sessions work for the president, and that the president has the power to fire the people who work for him. While Mr. Trump’s actions may be a crisis in a political sense, they are in no way a constitutional crisis or a blow to the rule of law or the American Republic.

    I think we can agree that a basic understanding of the US Constitution is a threshold requirement to be considered a serious political scientist.

    I note that Mr. Mounk is a lecturer on Political Theory. Perhaps he thinks the way to advance his career in the current academic environment is to signal his political bona fides by advancing such … shall we say … tenuous sentiments.

  6. isopar says:

    Mounk’s article is from 2016, why is this discussion happening now? It’s an interesting discussion, but what’s with the timing?

Leave a Reply