We're climbing hills and ignoring mountains

With any complex system, we can imagine the system as a function with a huge collection of input variables and a corresponding set of outputs. You can picture this as a vast landscape with lots of local maximums and minimums, each representing the result (measured using whatever output you want) of a given set of inputs, and a single global maximum. You can think of a system like the economy in this way, where a certain set of taxes, regulations, subsidies, supply and demand make up the inputs, and the output could be measured in a variety of ways like GDP per capita, average income, or some kind of loosely defined idea like general prosperity. The exact inputs and outputs don't really matter since we can't actually evaluate this function, but I think it's a good way to think about a complicated system.

When you imagine a system in this way, you can see that wherever we currently are in the landscape, the most obvious solution is to follow the gradient up the hill to a local maximum. And that's what people tend to want in real life. People fear change, so we like to try to improve our condition while moving as little as possible from the status quo. Take the health care industry for example. We can think of the inputs as number of doctors, number of hospitals, health care rules and regulations, subsidies, FDA rulings, etc and output could be measured as average cost per surgery, life expectancy, infant mortality, or something more general and loosely defined like overall healthiness to cost ratio. From where we currently sit on this landscape, it's easy to look around and see that we're not even at a local max. So naturally people want to climb the closest hill, towards something like universal state-run health care. It's the obvious solution from where we are, given the massive interference the government already has in the health care industry. But like most things, the obvious solution isn't always the best one.

We could evaluate our system of government in a similar manner. We could set the structure of government as the input, with variables like the relative power and authority of different government positions, the method of voting, campaign finance, the media's effect on voters, primaries, lobbyists, etc. The best way to measure output for this system is highly controversial, but you could imagine using personal freedom, economic equality, overall happiness or some other measure of well-being. For each of these methods of measuring the output, we would find a very different landscape generated from the same inputs. When you see the system in this way, it's easy to see how there can be such varied and yet strongly held opinions about which way our society needs to move. If you measure the system based on personal liberty, the gradient points in a completely different direction than on the economic equality landscape. But in either situation, the most common answer is to just climb the closest hill. Most people won't even consider a change in the larger structure of government, the variables that have the biggest effect on the landscape and the greatest potential to lead us to a global maximum. The vast majority only want to alter the relatively weak variables of which politician is currently in power, moving us slightly up the closest hill towards a short local max.

If we want the best outputs for any system, we have to take a step back and look at the entire landscape. We shouldn't be looking for the closest hill, we should be looking for the global maximum: that huge mountain of prosperity lurking far away from our current position. It's tough to see where it is when your stuck on the side of a hill. You have to challenge all the assumptions that drove you into that region of the landscape, because the mountain may very well be on the other side of the map. We must also lose our fear of traveling downhill, because we will have to cross countless valleys and small hills to get to the global max, but we should never loose sight of that mountain in the distance.