The Problem with Political History

As many readers know, my husband, Richard Stroup, died in November. For those who didn’t know him, here is a short obituary.  I very much appreciate the messages so many have shared with me about Rick.

Rick was somewhat skeptical of history as a discipline because he didn’t see any theory behind it (it seemed more like “one damned thing after another“). He preferred economic theory and its application to political behavior, which is called public choice economics. He and his coauthor James D. Gwartney were among the pioneers in this field.

Can history and economics be reconciled? Of course, but it’s not as easy as it seems. I have written about a leading effort to do that, cliometrics. But in this commentary I am asking a fundamental history question I often ask, “Can we predict the future based on the past?”

An article Rick wrote for the Concise Encyclopedia of Economics  may help answer the question. Economics is founded on the idea that people respond in a generally predictable way to personal incentives. This is fairly straightforward in the marketplace—each person tries to obtain what he or she wants by providing others what they want. Politics is different. Even though individuals still pursue their interests, the rules and consequences are different. Rick wrote:

“In democratic politics, rules typically give a majority coalition power over the entire society. . . . Political solutions can compel people, on threat of prison or worse, to financially support politically chosen goals.”

This coercion may be good (when correcting a wrong) or bad (when supporting a few at the expense of the many), Rick said. Whichever it is, a political decision can have a much bigger impact than a market transaction. Indeed, Rick’s article is a devastating litany of ways that political elections and bureaucratic actions help some and hurt others and open up wounds and resentments that play a part in subsequent elections. In contrast, a market transaction, when property rights are clear, is normally complete.

As I study political history (which most history is, I’m afraid) I am trying to apply that way of thinking. Perhaps you will detect it if you look here on my site  at the introductions (and links) to two stories I have published elsewhere. One is about the history of student debt and one is about the history of land-grant colleges. In both,  I try to untangle some of the unique factors that led to these two outcomes. Whether good or ill, neither of these, I believe, could have been predicted.

3 Replies to “The Problem with Political History”

  1. History and economics, like many of the sciences, now require interdisciplinary contributors if not team work. Example: the productive interplay of gene sequencing, ancient DNA, linguistics, and evolutionary psychology in determining human migrations, population dynamics. This new understanding, of course, raises many questions like has human evolution by natural selection stopped, and even if the interventions of civilization have weakened the species?

    Amid such complexity and for such a complex species we can predict only human behavior (which economists are pretty good at), but not the consequences. For instance, applying Kaufman’s first law of human powers, we can predict that all attempts to prohibit human gene editing and the creation of transhumans will fail. (The law is: no enhancement of human powers has ever been or can be successfully prohibited.)

    Perhaps I’m too late already in predicting that we will have interdisciplinary teams of scholars working with our best computing powers to both analyze the past and predict the future.

  2. “But in this commentary I am asking a fundamental history question I often ask, “Can we predict the future based on the past?””

    I think the answer to that question is “yes” or “no,” depending on the level of generality and levels of probability. This is not an exact science.

    For example, if a government seizes control of most of what was formerly a free market economy, we can say with a high degree of probability that the economy will stagnate or decline. We know this will happen within a relatively short time, but cannot know exactly the composition of the change. Some portions of the economy may continue to flourish as a result of unforeseen innovation and/or political favoritism.

    Similarly, if two hostile nations have approximately equal military capacity, we can say that they probably will remain at peace, but that if one of them lets its military position slide, the probability of war rises.

    On the other hand, we can predict likely war if a larger nation allows an aggressive smaller one to dictate its foreign policy responses. This factor may prove more weighty than the previous one, as happened just before before WWI: the Triple Entente and the Central Powers were in approximate equipoise, but Russia allowed its decision to go to war to be dictated by Serbia.

    My point is that history is full of lessons that can, in fact, allow us to predict the broad outlines of future events—or at least their relative probabilities.

  3. Jane, having been in the belly of the beast for 40 years, let me suggest that the best predictor of events is an assessment of the self-interest of those in power. Such assessments are both idiosyncratic and stochastic. (By stochastic, I mean randomly determined; having a random probability distribution or pattern that may be analyzed statistically but may not be predicted precisely.)

    Further, definition of “who is in power” is itself challenging.

    To suggest that a competent historian could model future behavior is something I think we need to leave to Hari Seldon https://en.wikipedia.org/wiki/Hari_Seldon.

Leave a Reply