Cassandra's DeLorean

Consider a policy that, in retrospect, proved to have pretty severe downsides. Now recall that there was probably someone who predicted this at the time, even if they didn’t get all the details right. Imagine if, at the time, this Cassandra was able to take the relevant elites (Congress, the Supreme Court, the Joint Chiefs of Staff, the board of directors for Consolidated Widget, whatever) on a fact-finding trip to the present day. These elites have a month to learn about how things turned out, whether by directly exploring or by spending a lot of time in the library. They only get the one trip and so they get to see how things actually turned out based on the policies that actually happened, but not the counterfactuals. At the end of the month, they return to the past where they can use their knowledge of the future to inform their own policy decisions. However the time machine is a military secret and so they are sworn to secrecy about what they saw. Now, here’s the question, would they act any differently?

I suspect that in some cases these policy elites would act differently, in other cases they would not, and imagining which is which is helpful for thinking about the limits of knowledge in shaping policy, or even action more generally.

  1. 1. This changes everything. We can imagine plenty of situations where knowing the future changes how we act. An obvious example is the Iraq War. If you know that Iraq had no active weapons program and that the Iraq War would take over a decade, kill almost 5000 uniformed coalition troops, not to mention lots of Iraqi dead (I say “lots” as estimates vary from the low to high end of 10^5 order of magnitude), and a stronger Iran. Likewise, my hunch is that Antiochus Epiphanes might have adjudicated the provincial power struggle between Jason and Menelaus a bit differently.
  1. 2. How do we know that caused the problem?. There are plenty of policies where causation is ambiguous enough that seeing how things turned out does not tell us definitively that it was caused by a particular policy. I am convinced that providing air support to the opposition in Libya and weapons and training to rebels in Syria not only prolonged bloodshed but led to a political crisis in Europe, but I also expect that if you crowded early 2011 Samantha Power, BHL, and Hillary Clinton into the DeLorean, they would not share my interpretation and they still would have done it. As a related point, the Republicans were never able to decide whether the problem with Obama’s Arab Spring era foreign policy was he did too much or did too little. Or let’s take domestic policy. Suppose you sent Lyndon Johnson a copy of Losing Ground on condition that he not show it to anyone. I can imagine that LBJ would simply say, “some of that stuff sounds pretty bad, but I don’t think it’s the fault of the Great Society.” I bet that a time traveler could talk Barack Obama into using different material at the 2011 White House Correspondent’s Dinner, but only because the upside was low so it doesn’t hurt to play it safe, not because the causal connection is clear.
  1. 3. Now we know this policy will cause problems, but we can’t not do this. Some actions are taken because they have forseeable bad consequences but are less bad than the alternative. This is basically Christensen’s model in Innovator’s Dilemma: an incumbent can know they are ignoring an emerging product category but still rationally stick to milking a declining but still profitable product category where they have comparative advantage. I think this probably also describes some actions taken during total war. If you told FDR abut the Cold War, he might have tightened up security at Los Alamos, but he still would have built the atomic bomb as long as you also told him about the battle of Okinawa. A more interesting category are actions that are committed because of a deontological commitment. Luther famously said “here I stand, I can do no other,” and so if you told him all about the Thirty Years War presumably he would keep on hammering his theses to the Wittenberg door.

Now, imagine that we don’t have a time machine, but we do have reasonable predictions about how policies will turn out. The difference between actually having a policy and just having predictions is it basically shifts most of #1 into #2, but it does nothing to change #3. So we can ask, how much would policy change if OMB had a DeLorean? Or if you prefer, what if George Soros, Tom Steyer, the Koch Brothers, or the Mercers had a DeLorean that they only lent to Congress strategically and selectively?