Acting Against Interest

February 28, 2012

I was reading a piece by a certain liberal columnist recently, and it struck me that the author, and other liberals I’ve heard, have two very different ways of thinking about voters’ acting against their (presumed, by liberalism) self-interest:

  • If they’re poor or middle-class, and they support lower taxes and less spending on “entitlement” programs (i.e., they want to seize less of the property of the wealthy for themselves), it’s treated as some kind of bizarre anomaly that cries out for explanation (possibly involving staggering stupidity), even a moral failing (e.g., by Paul Krugman or blog commenter Snoodickle).
  • If they’re rich, and they support higher taxes, they’re treated as having some kind of special authority to determine policy for the rest of us (including the rest of voters in their own income bracket), even moral virtue (e.g., by Paul Krugman and Snoodickle, and Snoodickle again).

I think the fact that liberals do not only one or the other but both gives the game away:  They’re not really adding anything epistemologically useful in either case by bringing in whether an actor is acting in his interest or against it.  Their position is for bigger government either way and regardless.

They might reply that my position is for smaller government either way and regardless.  That’s true; I favor liberty.  The difference is that I don’t try to play this game with argument from action against interest and try to have it both ways.

16 Responses to “Acting Against Interest”

  1. Matt Says:

    Excellent point!

  2. Snoodickle Says:

    Again, why do you have the moral authority to confiscate my property for the purpose of financing wars?

  3. Tevyeh Says:

    “I think the fact that liberals do not only one or the other but both gives the game away…”

    Be careful with your descriptions of what “they” do. You’ll sometimes find liberals (e.g. Krugman, Snoodickle) who express both of the views you highlighted above, and you may indeed have a valid “gotcha” with regard to these individuals. However, it’s fallacious to describe either of these views as “official” liberal positions. You’ll find plenty of logically or practically inconsistent views among the “members” of any political camp, with varying degrees of coherence.

    The attribution of an individual’s words, actions, or beliefs to “them” is probably one of the most widespread logical errors in political thought.

    • I think, based on my experience talking with various individual liberals or otherwise hearing from them, that the contradiction I described is part of a broader problem with how liberals think about things. But you’re right, I can’t point to particular, public examples, other than the two I just did, off the top of my head.

      • Snoodickle Says:

        e.g. Chillingworth

      • Tevyeh Says:

        It would probably be more meaningful to discuss these phenomena in terms of “a pattern I’ve observed in many liberals’ arguments.” An assessment of “how [members of a political/ideological camp] think about things” is necessarily subjective, and the person attempting it is at the mercy of his own biases. I won’t deny that there exist common intellectual or ideological threads that more-or-less “define” the various camps, but I’m not confident in my ability to fairly describe them.

        Check out for an example of a liberal (author ReflectionEphemeral) who’s always spouting absolute “truths” about conservatives and Republicans. He seems to have unshakable confidence in his complete understanding of “the Republican mind,” “the conservative ethos,” etc. He has a simple mental model of the American Right, a model with which he is comfortable and which he seems to cherish. He latches onto any instance of a conservative or Republican saying something asinine as further verification of his model (i.e. he’s a case study in confirmation bias.) Needless to say, his assertions are riddled with strawmen, bare assertions, and other logical fallacies.

        I much prefer to address individual arguments independently, trying to evaluate their logic and the assumptions upon which they are based. Once you start getting into general descriptions of the way “they” think, it becomes increasingly difficult to make an objectively meaningful statement.

      • Tevyeh Says:

        That said, I can think of a few tendencies that I generally associate with “liberals.” For example:

        —Conflation of positive and negative liberties, and a failure to appreciate the tension between them

        —Failure (or refusal) to look beyond the first (or first few)iterations of economic logic when attempting an economic analysis

        —The uncritical acceptance of the conclusions (or imagined conclusions) of empirical studies without regard for their quality or relevance, and their use as a “trump card” to conclusively “debunk” any “conservative” theory

        —Unearned smugness

        While I’ve seen numerous examples of each of these tendencies in people I’d describe as “liberals,” I’d never claim that they are universal within the American Left, nor that these are exclusively “liberal” tendencies.

      • Tevyeh Says:

        Woah…I just found a stark example of a liberal (Matt Yglesias) criticizing the exact behavior I identified as a common liberal tendency: “The uncritical acceptance of the conclusions (or imagined conclusions) of empirical studies without regard for their quality or relevance, and their use as a “trump card” to conclusively “debunk” any “conservative” theory.”


        Snoodickle—you’ve got to read this one.

      • Snoodickle Says:

        Here’s the converse problem. Conservatives use their hate of empirical studies, present company excluded for the most part, to at times disregard well accepted science. So there are problems on both sides.

      • Snoodickle Says:

        Also, I have always invited both you and Chillingworth to introduce into the discussion empirical studies that support your viewpoint as well. I am not guilty of using an empirical study as an alleged case closer without inviting competing studies.

      • Tevyeh Says:


        A few points:

        1) “Conservatives use…”

        That’s the kind of generalization I ranted about above. Your identification of a few exceptions (“…present company excluded for the most part…”) does little to mitigate the sweeping nature of your assertion. “Many conservatives use…” would probably be a fairer way to begin that sentence.

        On second reading, I notice that you included the words “at times,” and maybe this wording is meant to negate the generalization. If so, I still think a modification to the subject, “conservatives,” would have been clearer.

        2) “I have always invited both you and Chillingworth to introduce into the discussion empirical studies that support your viewpoint as well.”

        When the underlying phenomenon is the result of a compelex real-world process driven by a large number of variables, my consistent response has been that a simple two-variable plot or linear regression (the “analyses” we most frequently read about in newspapers and blogs) is usually worthless. You may as well seek answers from a magic 8-ball. If I dispute the conclusions of your 8-ball, please don’t ask me to counter you with the message from my own 8-ball. I don’t own one.

        I’m finishing up an MS in Quantitative Analysis. I won’t ask you to “trust” me on these issues, but I invite you to consider the possibility that I have some idea of what I’m talking about.

      • Snoodickle Says:

        I don’t understand how information can ever be entirely worthless, unless it is completely and utterly falsified. Information helps put theories into context. I accept that variables can account for a certain result in a study, and that the result, even if inconsistent with a given theory, does not necessarily disprove that theory. However, the information allows us to put that theory into context, and to consider the many variables that may have produced the result. Thus, we are thinking about and analyzing an issue based on both real world data and more abstract theories, and considering the variables that manipulate that data, and how those variables relate to the given theory. Anything short of that is just lazy.

      • Tevyeh Says:

        I should have been clearer. Information is never worthless, but shoddy analysis usually is.

        Too often, amateur analysts and political hacks try to measure the “effects” of variable X on variable Y by a simple plot or linear regression of Y on X. This approach only yields reliable results if you assume:

        1) that X is the only variable that significantly effects Y, or

        2) that all other variables that effect Y are held constant, or

        3) that all other variables that effect Y fluctuate randomly in a normal distribution around their respective means, with constant variance (homoskedasticity), and are independent.

        With regard to very many real-world variables (e.g. macroeconomic variables), these conditions almost never hold.

        Kevin Drum’s chart in the Yglesias post I linked to above is a perfect example of the kind of shoddy analysis I’m talking about. He attempts to “disprove” a causal link between tax policy and fertility rates by contrasting the graphs of “general fertility rate” and “total child tax subsidies” and noting that they don’t appear to show a strong correlation. He seems to completely ignore what I consider blindingly obvious: that fertility rates are impacted by variables other than “child tax subsidies.”

        Maybe a more rigorous analysis, incorporating other variables, would enable a more reasonable estimate of the effects of tax policy on fertility rates. However, the more variables you add to an analysis, the trickier it gets (i.e., to use an engineering analogy, it contains a lot of moving parts). For this reason, It may very well be extremely impractical to develop a tractable model capable of measuring this relationship with any degree of confidence.

        In this case (where a causal relationship cannot practically be identified and measured by competent empirical analysis), should we default to the assumption that such a causal relationship does not exist? I don’t think so. As a general rule of thumb, people respond to economic incentives. If our tax code provides significant incentives (or removes disincentives) for having children, I would have a hard time believing that such policy would have zero impact on fertility rates. I wouldn’t feel comfortable estimating the magnitude of this causal relationship, but I’d be confident saying that it’s positive.

        This kind of reasoning constantly informs our choices. I’m not aware of any rigorous empirical study that finds that locking your car doors is associated with a significant reduction in the probability of car theft—but I lock my car doors anyway. With regard to policy, we should make the best use of all available information, but we’ll have to face the reality that a good analysis will often be inconclusive. In these cases, I maintain that we should default to reason and logic. Human reason has its limits (real-world processes are sometimes counterintuitive), but I don’t think we should allow it to be trumped by shoddy empirical analysis.

      • Tevyeh Says:

        To be fair to Kevin Drum, I didn’t actually read his post or examine the study he cites. I’m relying on Matt Yglesias’ assessment. Shame on me—I hate when people do that.

  4. Snoodickle Says:

    P.S. I don’t mean to pick nits, but if one were to subscribe to the broader economic theory that the rich paying their fair share of the taxes is good for the economy, then both of the positions you have advanced above are completely consistent with one another. Gotcha!

Agree? Disagree? Thoughts?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: