Politics, US Election

Donald Trump and the Failure of Mainstream Social Science Part II

I published an article Sunday where I argued that mainstream social science is pervaded by ideology and that this blocks good scientific methodology. In this article, I cover a concrete example that is relatively recent and available online, a paper titled Trump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash.

Published in August 2016, the paper’s authors are Ronald F. Inglehart of the Institute for Social Research at the University of Michigan, Ann Arber and Pippa Norris, McGuire Lecturer in Comparative Politics at John F. Kennedy School of Government, Harvard University. It’s part of the Harvard Kennedy School Faculty Research Working Paper Series. Harvard is one of the world’s most prestigious public policy universities, and Norris is one of the world’s most cited political scientists.

The paper’s objective is to identify the main cause of the rise in support for ‘populist’ parties that have disrupted politics in many Western societies. Two causes are considered: economic disaffection and cultural disaffection. The paper examines each cause by connecting them to different demographics through hypothesis and survey results, and then comparing that to demographic data on supporters of these parties. They find that the main cause is cultural: religious people, white people, older people, and men are overrepresented among voters of populist parties, and since these groups tend to hold traditional values they are most likely disaffected by cultural change.

Superficially, this sounds like good scientific methodology. I will argue that it is not, due to the insertion of ideology that blocks genuine scientific analysis, and that the conclusions of the research suffer from it.

The first methodological problem is the pervasive use of morally charged language. For example, the words xenophobia/xenophobic appear 15 times in the paper. One instance is part of the definition of their object of analysis (p. 7):

“Populism” is a standard way of referring to this syndrome; emphasizing its allegedly broad roots in ordinary people; it might equally well be described as xenophobic authoritarianism.

The words authoritarian/authoritarianism appear 21 times (including figures). For instance (p. 6):

During the last two decades, in many countries, parties led by authoritarian leaders have grown in popularity, gaining legislative seats, reaching ministerial office, and holding the balance of power.

The words nativist/nativism appear 6 times. For instance (p. 30):

Populist forces have already proven decisive for the outcome of the British referendum on membership in the European Union in June 2016, with their leaders igniting anti-immigrant and nativist sentiments in England.

The problem with morally charged words is that they are effectively a conflation of two things: a descriptive statement and a moral judgement. This leads to two problems.

First, they give the appearance of being scientific without really being scientific. The word xenophobia uses the naming practice commonly used to diagnose psychological disorders — phobia — to make it seem like questioning large-scale immigration or critiquing cultural practices is a psychological disorder. It isn’t. It’s a moral judgement designed to seem scientific, especially to non-scientists who find themselves browbeaten by words like this. The danger is that scientists themselves are tricked into thinking their analysis is more objective than it is.

Second, it inserts vagueness into the analysis because moral beliefs themselves are vague and often implicit. Is it nativist to cheer for the national football team? Is Barack Obama authoritarian? Are people who criticise female circumcision xenophobic? There are no accurate answers to these questions since the words are based on vague moral judgements. What ‘xenophobe’ effectively means is ‘a person who criticises foreign cultures or resists immigration to the extent that I personally consider immoral’.

This doesn’t just apply to negatively charged moral language, but to positively charged moral language as well. For example, the word progressive appears in the paper 28 times. For instance (p. 14):

In particular, it is well-established that education, age, and gender are strong predictors of support for progressive values.

The effects are the same as with the negatively charged moral language: introduction of vague, moral beliefs into the analysis are made to seem scientific. The word progressive relies on the notion of an arrow of time stretching from the past into the future, and ties a set of values to that notion. But, of course, it’s not a given that these values will stretch endlessly into the future. After all, many Christians were convinced when Christianity was the dominant belief system that the spread of Christianity was identical to progress, and that it was destined to spread everywhere eventually.

Furthermore, it establishes through assumption that since the spread of a particular set of values is identical to progress, anyone opposing these values is backward by definition.

All this morally-charged language introduces vague moral ideology into the analysis, under the guise of science. The consequence of this reveals itself when the authors develop their analytical framework and end up with a framework that seemingly aligns perfectly with their ideology. (At least with respect to cultural values, which is the core of the analysis.) This is unsurprising; when you articulate your research in morally charged language, you inevitably end up with a moral framework.

The framework developed by the researchers is essentially a continuum from moral to immoral. On the one end on the continuum sits an amalgamation of all that is moral to the authors: pluralistic democracy, tolerant multiculturalism, multilateralism, and progressive values. These people are cosmopolitans, which as the authors point out is Greek for ‘citizen of the world’. On the other end sits an amalgamation of all that is immoral to the authors: anti-establishment sentiment, a strong leader, popular will, nationalism and traditional values. These people are populists, or as the authors point out earlier in their paper, they could equally be described as xenophobic authoritarians.

This is not an empirically sound framework, it’s just a bunch of vague, morally charged words. Needless to say, problems arise when the authors then try to use this framework for their empirical analysis, which amounts to placing voters somewhere on the moral-to-immoral continuum.

Firstly, it fails to account for conflicts within these values. They’re so loosely defined that it seems like they all fit together seamlessly, until they are applied to real-life situations. Consider the authors’ take on the resistance to political correctness that to various degrees characterises much recent political debate (p. 29):

Less educated and older citizens, especially white men, who were once the privileged majority culture in Western societies, resent being told that traditional values are ‘politically incorrect’ if they have come to feel that they are being marginalized within their own countries.

And, with reference to Donald Trump (p. 31):

His rejection of “political correctness” seems particularly appealing to older, religious white traditionalists who find themselves left behind by growing support in America for such issues as same-sex marriage, rights for transsexuals, gender equality for women in politics, and immigration rights.

This follows automatically from the authors’ framework. They’ve effectively locked in a set of values under the implicit label of ‘the Good’, and so they’re forced to rationalise opposition to any of it as being immoral (i.e., ‘backward’). There is no other option, because there’s nowhere in the framework to place people who, for example, think that the overzealous enforcement of multiculturalism and/or sexual liberties has stifled free speech on topics such as religiously-motivated terrorism.

And secondly, there is nowhere to place people who think that the large waves of Muslim immigrants in Europe has suppressed cosmopolitan values. The irony, of course, is that many of these immigrants tend to be more religious and less tolerant of various rights than even people that vote for populists. Are these immigrants xenophobic? Are they authoritarian? And where do they fit in the framework?

So there’s nowhere on the moral continuum to place people who resist various forms of cosmopolitan values because they think they conflict with other cosmopolitan values. Nor is there anywhere to place the immigrants, who in many cases would be far toward the ‘immoral’ pole on the authors’ continuum, except that they don’t fit the narrative of the backward old white Christian man.

And if one were to look beyond ideology, one would likely find that these two factors are a huge driver of populist support. The notion that many people in Europe are resisting large-scale immigration because they’re clinging to traditional values is ridiculously simplistic.

But the authors can’t possibly know this, because their analysis never has a chance to get off the ground. By introducing their ideology through morally charged language that then coalesces into a moral continuum, their framework, and thus their analysis, lacks the degrees of freedom to separate the underlying variables where they conflict. And so, unsurprisingly, they end up with the conclusion that their ideology is correct.

And this affects the predictability of their model. The only prediction the authors explicitly make is that populism is likely to continue, which isn’t much of a prediction at all. Implicitly, though, their conclusions point to a prediction that as time passes and older generations of white people die off, Western society will become more ‘cosmopolitan’, i.e., with more pluralistic democracy, tolerant multiculturalism, multilateralism, and progressive values. But is this true?

There are two factors that suggest it might not be the case. Firstly, the fact that the main source of population growth in many Western countries are immigrants or second-generation immigrants who tend to be quite religious and not actually that tolerant of the authors’ cosmopolitan values. Secondly, there appears to be a conflict between certain elements within these values, for example between rights-enforcement and free speech, which might challenge the notions of pluralistic democracy. But these issues are not addressed at all, because they are made impossible a priori by the authors’ ideology.

 

Uri Harris has an M.Sc. in business and economics. His work can be found at theamoralsociety.blogspot.com.

Photo by Gage Skidmore.

 

See also: Donald Trump and the Failure of Mainstream Social Science Part I

Donald Trump and the Failure of Mainstream Social Science Part III

29 Comments

  1. benjamin winegard says

    I really couldn’t disagree with this critique more. Welzel, Inglehart, and Norris have been conducting cutting edge political/cultural analysis for decades and they have a coherent framework that is both descriptive and predictive. I would suggest reading their longer expositions (especially Welzel’s “Freedom Rising” which is a brilliant book).

    Specific points:

    1) “The framework developed by the researchers is essentially a continuum from moral to immoral. On the one end on the continuum sits an amalgamation of all that is moral to the authors: pluralistic democracy, tolerant multiculturalism, multilateralism, and progressive values. These people are cosmopolitans, which as the authors point out is Greek for ‘citizen of the world’. On the other end sits an amalgamation of all that is immoral to the authors: anti-establishment sentiment, a strong leader, popular will, nationalism and traditional values. These people are populists, or as the authors point out earlier in their paper, they could equally be described as xenophobic authoritarians.

    This is not an empirically sound framework, it’s just a bunch of vague, morally charged words. Needless to say, problems arise when the authors then try to use this framework for their empirical analysis, which amounts to placing voters somewhere on the moral-to-immoral continuum.”

    Meh. No. The words are meaningful enough for social scientists to understand them and are defined rigorously in other writings. These are standard social science terms with operationally clear definitions. You are free to use your own argot if you think these terms are morally charged but they are not inherently moral. In fact, many right-wing parties believe that tolerance and multiculturalism ARE morally bad in many circumstances. Inglehart, Welzel, and Norris use valid and reliable scales to measure these constructs and have been collecting data for decades. You can look at the factor structure and other psychometric properties of these scales if you desire.

    And it’s just a flat untruth to state that the authors place voters on a “moral-to-immoral continuum.” That requires some heroic hermeneutic gymnastics to pull out of their writings!

    2) “And secondly, there is nowhere to place people who think that the large waves of Muslim immigrants in Europe has suppressed cosmopolitan values. The irony, of course, is that many of these immigrants tend to be more religious and less tolerant of various rights than even people that vote for populists. Are these immigrants xenophobic? Are they authoritarian? And where do they fit in the framework?”

    Really? These same authors have written at length about the relative values of different cultural/religious groups. And, yes, they have demonstrated that Islamic societies, and individuals, in particular, score low on secular and self-expression values. Again, I would suggest reading other scholarship produced by these scholars before making false assumptions.

    Here’s just one passage from Norris and Inglehart about Islamic societies:

    “Muslim societies are neither uniquely nor monolithically low on tolerance toward sexual orientation and gender equality. Many of the Soviet successor states rank as low as most Muslim societies. However, on the whole, Muslim countries not only lag behind the West but behind all other societies as well. Perhaps more significant, the figures reveal the gap between the West and Islam is even wider among younger age groups. This pattern suggests that the younger generations in Western societies have become progressively more egalitarian than their elders, but the younger generations in Muslim societies have remained almost as traditional as their parents and grandparents, producing an expanding cultural gap.”

    The entire article, very worth reading, is available here: http://foreignpolicy.com/2009/11/04/the-true-clash-of-civilizations/

    3) “Secondly, there appears to be a conflict between certain elements within these values, for example between rights-enforcement and free speech, which might challenge the notions of pluralistic democracy. But these issues are not addressed at all, because they are made impossible a priori by the authors’ ideology.”

    This, again, is pure fiction. Christopher Welzel, in particular, has talked about liberal illiberalism and emancipative values. Inglehart and Norris have also talked about some interesting contradictions that occur in societies that score high on self-expression or emancipative values.

    • Uri Harris says

      Thanks for your comment.

      A couple of responses:

      1. If their framework is as predictive as you say, why does it seemingly completely miss the debate on political correctness that has been going on, and which Jonathan Haidt, Sam Harris and many others have covered? Or do you think these people are simply clinging to their traditional values? It’s not the authors disagree, they don’t even acknowledge it (see the quotes above). Contrast that with my suggestion, that their framework makes it effectively imposible for them to seperate critique of free speech supression from ‘xenophobia’.

      2. It doesn’t take hermeneutic gymnastics at all. Do you disagree that their cosmopolitan category and its contents are positively morally charged, and the populist category the opposite?

      • benjamin winegard says

        I would read “Freedom Rising” before making these statements. Both are clearly addressed in that book.

    • Jon Taylor says

      Your argument is mostly a hyperbolic appeal to authority.

      Try a thought experiment. Imagine if a conservative had written the description for the “populists” and the “cosmopolitans”. It might go something like this:

      “Populists believe in equality before the law (one vote for one person), they respect that democratic elections are a civilized way of resolving disagreements between political factions and trust in the wisdom of the crowds. They are epistemologically humble and thus trust traditional norms and rules to guide their behaviour and are suspicious of faddish trends. They place faith in tried and tested social units such as the family and nation-state. They are loyal and have respect for hierarchy and authority figures. They like to have dogs as pets.”

      On the other hand –

      “Cosmopolitans tend to devalue the opinions of those who are less intelligent or educated as them. They believe in technocracy and in the right of supra-national corporations and bureacracies to usurp local laws and customs. Cosmopolitans, due to their education, are in the best position to socially engineer society, if and when engineering needs to take place. They are morally virtuous and are therefore in a legitimate position to shame those who don’t conform to their values. Cosmopolitans do not believe that the nation-state is a desirable unit of social organization, and would prefer to live in a borderless, stateless global society. They believe the family is a historical site if oppression for women and LGBTQ people and thus would like it to be eventually phased out. They like cats as pets”.

      I’ve actually tried to be neutral here, but it is really easy to cast cosmopolitans as a bunch of jerks and populists as nice people. It depends on your bias. Arguing that the authors of the paper don’t have a bias probably only betrays your own. They repeatedly use emotionally charged descriptors in the paper, e.g. “angry”, “paranoid” etc. It’s fairly obvious that they have a bias.

      • benjamin winegard says

        You are free to operationalize your terms as you like. I’d suggest reading Inglehart’s and Welzel’s broader work before casting aspersions at them.

    • patricia says

      @benjamin what’s with the self-righteousness? This article is basically just saying the same thing as this: http://www.rci.rutgers.edu/~jussim/Duarte%20et%20al,%202015,%20BBS,%20target,%20commentaries,%20reply.pdf

      instead of arguing that there should be no ideological diversity he is saying that there should be no ideology. Surely the ideology problem is even worse in political science than in psychology? Or do you think that there are no problems of ideology in academia?

      • benjamin winegard says

        Not sure this how exactly this is relevant to Inglehart, Welzel, & Norris.

        • patricia says

          Inglehart, Welzel, & Norris are academics, right?

          Academia, and the social sciences has a problem with ideological conformity right?

          I read the first 7 pages of their article on UKIP and Trump and it positively drips with ideological bias. I couldn’t read anymore.

          I am sorry someone came along and criticised your favorite academics. These things can be triggering, we understand. Hopefully your university has a safe space you can go to.

    • Benjamin Winegard:

      “I really couldn’t disagree with this critique more. Welzel, Inglehart, and Norris have been conducting cutting edge political/cultural analysis for decades.”

      I thought at first your comment was intended to be satire. But apparently it isn’t. You disagree with Uri because these people are well established experts in the field! This is exactly one of the main problems with academia in general – appeal to authority, too much respect for so-called experts. I

      This also shows that the authors of this paper have got their use of the term “authoritarian” backwards. It’s the establishment elite that populism is challenging who are the true authoritarians, not the populist movements. But because ‘authoritarian’ sounds like a bad thing, the authors turn it around.

      Then you suggest that Uri should read one of their longer expositions. Good grief. The poor guy has already waded through 53 pages of sociological waffle and you expect him to read something even longer?

      Finally, you are wrong when you claim that the authors are using precise terminology without any moral charge. Did you notice that the authors apply the term “rabid” in the context of UKIP? Is that a clearly defined term without any moral implications?

      • benjamin winegard says

        I’m not appealing to authority. If my argument were “Ronald Inglehart can’t be wrong because he’s an expert” that would be an appeal to authority.

        • So we have four short responses from Dr Winegard. None of them addresses any of the issues raised. Two say “go and read this” while the last denies appealing to authority.

          This is disappointing, since he himself has written at least one interesting paper about “the dangers of ideological uniformity in the social sciences” citing Tetlock, Jussim, Haidt et al,

        • patricia says

          “I’m not appealing to authority”

          hahahahahahaha!

          From your first comment

          “I really couldn’t disagree with this critique more. Welzel, Inglehart, and Norris have been conducting cutting edge political/cultural analysis for decades”

          If you don’t recognize that as an appeal to authority you need some lessons in basic logic.

        • Benjamin you seem to be missing the part when you are appealing to authority and when you make a cast iron statement about it. I think you are hoisted by your own post. I also object to replies that say, go read this book or this paper without any synopsis of the suggested paper/book, and is often used to shut down debate. This tactic is used extensively by climate scientists attempting to hammer home a soft point. It is childish and to my mind pathetic.

    • Tim Hammond says

      Your claim are little more than a fallacious appeal to authority – They are well-regarded authors, these supposedly moral words are used by sociologists etc”

      The simple fact is that this piece of analysis uses words like “resent”, which are wholly moral and wholly assumed. And once I “resent” something, it is not longer a rational, thought-through belief, but an irrational, emotional one. And there you have the claims in a nutshell – those who supported Trump and/or Brexit are emotional non-thinkers, those who did not are cool, rational and well-informed.

      You can dress those claims up with models and stats all you like, but they stay the same. And the stay self-congratulatory, self-righteous claims driven by prejudice.

  2. Jon Taylor says

    My comment above was meant to a reply to Benjamin Winegard’s comment.

  3. DiscoveredJoys says

    “Superficially, this sounds like good scientific methodology. I will argue that it is not, due to the insertion of ideology that blocks genuine scientific analysis, and that the conclusions of the research suffer from it.”

    Actually, and I’m not trying to be argumentative (I’ve not yet read the paper), it struck me that the authors are not doing ‘science’ but ‘philosophy’. Philosophers tend to start off with an assumption and then derive the consequences from it, thus hoping to set out ‘how to live’.

    If it *is* science then you would need empirical evidence or observation collected (before and after) without reference to the underlying assumption, in this case some version of ‘Liberal Values’. Otherwise you are just explaining why the world is (or should be) pink whilst wearing rose-tinted glasses.

    • Uri Harris says

      Thanks for your comment.

      I didn’t make it clear enough. The authors *do* derive their inputs from empirical studies. They’ve analysed trends in different societies and identified patterns, for example what characterises democratic, Western societies. They’ve also studies trends within these societies, which point to consistent developments over time. I didn’t mean to imply, of course, that they’ve just made up a moral framework. My point is that by interjecting morally charged language their framework *becomes* a moral framework, thus making it more difficult for them to address new trends. Hence they are essentially forced to assume that challenges to political correctness are due to people clinging to traditional values, for example.

      • DiscoveredJoys says

        Thanks for replying. Perhaps my original comment needs clarification too.

        I have no reason to doubt that the authors derive their input from a selection of empirical studies. Was that selection unbiased? Did the authors use morally charged language because they interpreted the results of studies through an implicit moral language framework?

        • Uri Harris says

          Yes, you hit the nail on the head. I think that pervasive use of moral languange (among other things), prevents them from recognising problems in their framework (their framework is part empirical and part moral/ideological, and that’s a problem).

  4. Whyaxye says

    Another excellent article. Just from the excerpts alone, one can see that this “study” is an ideological polemic disguised as a piece of “social science”. I would guess that by appealing to “science”, the authors and others like them are merely aiming for a sympathetic hearing and acceptance. The same happened when religion was less questionable: minds were swayed by an appeal to Christianity. What reasonable person could demur if something were presented as the Christian option?

    This bit is particularly good:

    “By introducing their ideology through morally charged language that then coalesces into a moral continuum, their framework, and thus their analysis, lacks the degrees of freedom to separate the underlying variables where they conflict. And so, unsurprisingly, they end up with the conclusion that their ideology is correct”

    I think good practice would be for researchers (or polemicists who claim to be doing research) to state clearly at the start of their paper some clear definitions of their key variables. What exactly do they mean by “progressive”, “authoritarian”, “liberal”, and so on. These are much more like subjective opinions (albeit shared) than decided and settled definitions. If physicists were as sloppy about terms like “mass”, “velocity”, “energy”, etc., the paper would be deservedly risible.

    The authors having done that, the reader could then decide whether the definitions bore any relation to reality. What makes a movement “authoritarian”? Do we agree with the author(s)? Thereafter, we could view the conclusions of the research in the light of those definitions. “OK, there is a rise of authoritarian parties in Europe, but only in the weak or daft sense that the authors are using the term.”

    • Uri Harris says

      I agree. They do have definitions for many of those terms, but I think that by using morally charged words they also build in an ideology that makes it more difficult for them to recognise new trends. See my latest article.

  5. At the risk of provoking hysteria, let me point out that this is exactly what happens in what currently passes for mainstream climate science. The prevailing narrative has taken on a moral framework which automatically demonises and devalues research that undermines its story.

    And the ‘we are all going to die unless governments save us from ourselves’ attitude hooks in neatly to the liberal agenda which Uri has described as ‘cosmopolitan’ — essentially an ongoing and endless transfer of power and autonomy from the individual to the state.

    • patricia says

      I am pro-climate-change-intervention from the point of view of the precautionary principle but I hear what you are saying.

      This type of narrative formation cloaked in scientific nomenclature ultimately becomes unfalsifiable.

    • Jon – I totally agree and it is the smell of ideology that makes many sceptical scientists think something is on the nose, this makes them look closer. Then you get comments from people like Richard A. Muller (uni california – berkley earth project) referring to the ‘hockey stick’ as giving him a list of scientists who’s papers he would not read. He believes climate change is real.

    • Uri Harris says

      Exactly. Insisting on linking science to a particular set of moral beliefs makes people sceptical of the science.

  6. Karmakin says

    Very good article. What I personally see the problem is, as described here really is a lack of understanding of the political landscape as a whole. It indicates a very real clinging to a concept of political ideology as a binary, or at best a binary spectrum, you have your “Progressives” and your “Conservatives”…we’re running to a point where increasingly that’s breaking down. As said, we’re being forced to choose which Progressive ideals to support, as they’re becoming increasingly contradictory.

    Generally speaking, traditional social science models simply can’t handle this very well, and it’s leaking into society as a whole.

    • Uri Harris says

      Thanks you. I agree, there is aneed for a better framework. Also agree that it does leak into society and helps create a tribalist mentality among people.

  7. Uri, did you contact Inglehart and Norris to make them aware of your public criticism of their paper? I think it would be courteous to do so and it might be interesting to see their response.

    • Uri Harris says

      No, I didn’t. I suspect writing to them would be a waste of both my time and theirs, although to be honest I hadn’t even thought about it until I read your comment.

Comments are closed.