Hacker News Reader

Comment by user3939382 | Hacker News Reader

It's interesting that many of us here are programmers and understand that debugging becomes exponentially more difficult as you add a number of variables beyond 1 to your deductive tests.

Yet when it comes to the economy, which is the product of an unfathomable number of variables, from the same people there are consistently these confident assessments of why we see x, y, z and what will happen.

It's worth discussing those things to the best of our ability, but we should have a tone of our confidence in these types of assessments that's commensurate with reality.

View on HN

I've always thought that the Fed gets too much blame for the economy going badly, when all they can control is the interest rate and QE/QT. It's like an inverted pendulum PID problem[0] but with an unfathomable number of variables and stakes.

0. https://www.youtube.com/watch?v=I5GvwWKkBmg

They introduce top down decisions on what is a bottom up complex emergent system. Beyond being an exercise in futility to control it that way, it introduces irrationality to the whole thing as it becomes a politically driven thing. That irrationality makes everything else behave that way and we loose any sense of "natural" trends and predictability. The bailouts are a perfect example, we are now surrounded by zombie companies and uninteted consequences which are making the whole thing more fragile and unstable.

Monetary policy has incredible sway over the economy, so they should bear the brunt of the blame. It's odd to me that everyone talks about how important democracy and elections are, but when it comes to the mighty dollar, the thing that runs this country and largely the world, all decisions affecting its value are made by a board of unelected individuals behind closed doors through opaque models. The FOMC membership itself partly consists of bankers elected by other bankers, which always seemed like a conflict of interest to me. The Fed deserves more scrutiny, if anything.

Right, but it's still possible to understand and debug incredibly complex programs, in the end it's still deterministic.

The problem with economics is that it's a social science, and that humans are non deterministic in their behavior. Not even seasoned economists can predict an outcome, the best they can do is understand what happened after the fact.

Separate to the "(non)-deterministic" question which would detract from this discussion.

One part we also fail to consider is that there is a feedback loop between all the individuals in society and any large economic levers that get pulled. This along with all the complicated interactions that are a result of the specific imperfect set of information each individual has makes this problem next to impossible to fully model. At best we could have approximations and models that abstract that complexity away. We approach something similar to the N-body problem in physics, which on some level I'd argue is simpler even as we have specific equations government their behavior. Not so the case for humans and their "internal" decision making process.

In my opinion it's classic Dunning-Kruger effect. For some reason the people in question took some Econ 101 or read some thing somewhere, and are extremely confident in their extremely limited economics knowledge, being completely ignorant of how complex it actually is in the real world.

As an aside, if there's one thing strategy games (and in particular Paradox' Grand Strategy games and Democracy) have taught me that might actually be real world applicable, is that nothing is simple or easy, and anyone (especially politicians) that they have one easy solution that will fix all problems is either ignorant or a charlatan.

Reply on news.ycombinator.com