Thursday, March 31, 2016

Preventing primary disasters

Forget who the candidates are for a moment. Everything that can be said about them has been said. (Some things bear repeating, but I'll leave that to others.) I'd like to focus on how we got to where we are, and how we can avoid such messes in the future.

Right now the leading Republican candidate has 48% of the delegates and 37% of the votes. At this point, no matter who wins, the party nominee will be someone that two thirds of Republicans voted against.

This is bad policy from every possible perspective. On the level of principles, it's just undemocratic. On a strategic level, it depresses voter turnout in the general, because most Republican voters will feel robbed. Justifiably so!

This has nothing to do with who the candidates are; we could see the same outcome with an entirely different set of candidates. It has everything to do with the systems the state parties have put in place.

There are two problems that need to be addressed. First, how delegates are allocated.

Ohio and Florida are absolutely crucial in the general election. The Republican party winning the White House is largely contingent on voter turnout in these two states. But both primaries are winner-take-all, and in both, the state was won by someone with less than half the votes. Over half of Republican voters in Ohio and Florida have been stripped of any voice in selecting their nominee. In South Carolina, that number is closer to two thirds.

Pretend you're one of the voters whose vote was thrown out by this system. Are you more likely to show up in November? Or less?

States using semi-proportional systems also contribute to the problem. In Alabama, a candidate with 21% of the vote got 13 delegates; a candidate with 19% of the vote got one delegate. In what universe is this giving each voter anything like equal weight? Winner-take-all is a huge problem, but winner-take-more isn't the solution.

The state parties should all adopt straight proportional allocation of delegates. This would at least minimize the disparity between popular vote and delegate count, and give all Republican voters an equal voice.

But this doesn't solve the more fundamental problem: a candidate with a third of the vote would still be winning. The reason for this comes down to two words every candidacy dreads: vote splitting.

The way we cast votes in this country breaks if there are more than two candidates. We all know how this works: where one candidate running alone might win easily, if there's a similar candidate on the ballot, they split votes between them, and the least popular candidate ends up winning. That's why we have exactly two major parties, both of whom dread a strong third-party run. That's how HW Bush lost to Clinton, and how W Bush won over Gore. And that's why there have been constant calls during this primary season for candidates to drop out early.

If there are more than two candidates, everything goes to hell.

This is directly caused by the way we cast votes. There are two or three or ten candidates on the ballot; you vote for one, and by extension, against all the others. This system is sometimes called plurality voting, or first-past-the-post voting. I like to call it by a more direct name: pick-one voting. Sure, there are other methods of voting. But this is America, and that's just how we do things here, right?

Well, no.

Pick-one voting is nowhere in the US Constitution. It's nowhere in any state constitution or law I've ever seen. None of our founders ever sat down and wrote, "Out of all the possible voting systems, pick-one is best, and here's why." Nobody decided to use pick-one voting. We vote this way because we always have. Because of it, we end up selecting a standard-bearer who commands a solid minority base, but who the majority can not support.

This is no way to run any organization. But the state parties can fix it.

There are many other voting systems out there. A few cities use instant-runoff. Others will extol the virtues of the Condorcet methods, or range voting, and they have valid points. But for real-world elections, the best system by a mile is approval voting, because it's simplest to understand, and trivial to implement. No money need be spent; it requires no new voting machines, because all machines already support it. Votes can be counted exactly as they are now. It's even simpler than pick-one, hard as that may be to imagine!

The only difference with approval voting is that the voter now marks every candidate they approve of. Your vote is now a "yes" or a "no" to every candidate, instead of being forced to vote "no" on all but one. Want to cast a vote for "anybody but him/her"? You can do that! Want to vote for a non-establishment candidate, but you don't care which one? Not a problem! And since every voter gets to vote "yes" or "no" on every candidate, every voter still gets an equal voice.

This primary season has been a disaster for the Republican party. No matter who wins, the party is more divided than at any time in living memory. Think how this primary election would have gone under approval voting. We could have started with the same candidates, but instead of ending up with three that represent three disparate wings of the party, we would have ended up with one who the whole party could support.

I have no idea at all who that would have been. But this I know with all certainty: the Republican party would have come out stronger and more unified.

Saturday, March 19, 2016

Working for a living

Let us assume most people have a goal of not dying. People will take whatever steps they deem necessary to accomplish this goal.
People need certain things to survive: food, water, shelter, clothing, energy, medical care. They also need secondary things that help us get their primary needs: education, transport, communication. Lots of thing-needing going on.

We live in at a pretty nice point in history. The things people need to survive exist in the world! This is much better than the alternative, where we have to actually create what we need. (Admittedly some people still do that, but they're outside the scope of this article.) Since our survival needs are owned by someone, survival is simply a question of effecting the transfer of these things. There are three basic ways of doing this:
1) Steal. Someone gives you what you need involuntarily. This could be burglary, fraud, looting, war, or other forms of theft.
2) Trade. Someone gives you what you need voluntarily, in exchange for something else they want. This includes shopping, barter, and labor.
3) Gift. Someone gives you what you need voluntarily, with nothing in exchange. This includes charity and government welfare, though government welfare includes some involuntary aspects.

All people do some combination of these three things to achieve continued survival.

First, we want to avoid theft, if nothing else because it's destabilizing to society. If theft is endemic, it forces everyone who isn't stealing to spend additional resources on security, which are resources spent not making the world better. We try to dis-incentivize theft through punishment. But no matter what punishment scheme you put in place, people will steal if they can't get what they need to survive, either through trade or gift. Thus, to prevent theft and maximize overall efficiency, we must make sure that survival is otherwise achievable through trade or gifts.

First, we consider trade. The overwhelming majority of humans have exactly one thing to trade: labor. We work, and in exchange we obtain things we need. Our labor has a value, and our survival needs have a value. The ratio of these determines how many hours one must work to survive. This number will vary from person to person, year to year, and place to place.

But what if that ratio gets out of whack? What if the typical person needs to work fifty hours to survive? Seventy? Ninety? If the number of labor-hours required to survive is more than a typical person can supply, if the cost of living goes up or the value of labor goes down too far, it becomes literally impossible to work for a living.

(Now, I am not presently arguing that this has occurred today, here, or at any other time or place. I am simply pointing out the boundary conditions of our present system.)

Supposing this occurs, and that we still want people to not turn to theft, we have a few options:
1) Make labor worth more than its market value. This can be done with minimum wage laws, or with the artificial creation of new jobs.
2) Make cost of living less than its market value. This can be done with price controls, or with subsidization of survival needs.
3) Decouple cost of living from labor. Give people their survival needs whether they've earned them or not.

None of these are free-market solutions; all are some form of government intervention. From this, we have an inescapable conclusion: the free market only leads to a stable society as long as people can earn a living. Once the cost of living exceeds the labor available to an individual, for whatever reason, government intervention is required to preserve stability.

Of course, that leaves trivial details like when and how...

Saturday, March 12, 2016

The name of God

Read the Old Testament, and you see God doing a lot of things that, to us, may make little sense. But God had an intended audience, and it wasn't us. You have to ask, who is He doing these things to and for? He's doing them for people in a world full of gods. And he's doing them to distinguish himself from those gods. What lessons would a people in that context learn?

Noah. Noah and his family are the only survivors of the flood. They learn that there is behavior God will not tolerate forever. But they also learn that God saves those who obey him.

Job. What does Job believe about God when the story starts? Read chapter 9; Job believes God is treating him unfairly, and wishes for a mediator, but there is not one. Yet in chapter 19 Job says he believes that his redeemer lives, and will come, and that he will have his day in court. Job serves his god, but he also believes in some power beyond that god. At the end, YHWH comes, and tells Job he has it wrong: the god he serves is the god of all creation, and that there is no appeal.

Abraham is told by YHWH, "do what I say and you will be blessed." Abraham lives in a world full of gods demanding human sacrifice. So when Abraham has a child in his old age, and YHWH tells him to sacrifice that child, Abraham is willing. But then YHWH stops him. Now Abraham knows that this god is not like the others; He wants obedience, but human sacrifice is off the table. And when He makes a promise, that promise is kept.

Moses is told "I do these things to make a name for myself. Tell them I AM sent you." God's name is his reputation. He wants to be known throughout the region as "I AM", the god that exists, as opposed to the ones that don't. And through the remainder of the Old Testament, we see God identifying himself in terms of his existing reputation: I am YHWH, the god of Abraham, Isaac, and Jacob, who led you out of Egypt.

Story by story, generation by generation, God built a reputation as a god who exists, who is powerful, who is merciful, who keeps promises, and who punishes evil. And that reputation is referenced again and again throughout the prophets and the New Testament.

Wednesday, March 9, 2016

Open letter to the Tennessee Republican Party


Put aside the details of the candidates for a moment; the primary elections have a serious structural problem that needs to be fixed before 2020. If the Republican convention were held today, the leading candidate would have received only 34% of the votes. Put another way, two thirds of Republican voters would have voted for someone other than the winner. How can this result possibly unite any party?

This comes from a basic flaw in the most common voting system we use: there are three or five or ten candidates, and the voter picks exactly one. Essentially, the voter says “yes” to one candidate, and “no” to every other. This system only works if there are exactly two candidates. Otherwise you get vote splitting and division, exactly like the Republican party is seeing now. This is fundamentally why we have two parties and primaries to start with. But now the primaries themselves are seeing vote splitting.

It would be vastly better for the party to use approval voting: there are three or five or ten candidates, and the voter marks as many as they find acceptable. Put another way, your vote is a “yes” or a “no” on each candidate. It’s easy to understand, all voting machines already support this process, and few if any laws need to be changed. The party could have had a much more unifying and successful primary season with this approach. I hope that by 2020 the Tennessee Republican Party will adopt approval voting.

Saturday, March 5, 2016

Dollars per life and the TSA

I've written previously about how much money we spent in Iraq, and what other things we could have done with that sum of money. I've also suggested that taxes should be seen as transactional; are you getting what you pay for? I'd like to focus that discussion. Let's look at causes of death in the United States, and what we spend preventing them. In particular, let's look at terrorism, and the TSA.

TSA was created in response to 9/11, to prevent other similar attacks. Since then, TSA has been budgeted something like $7 billion per year. Assume they've prevented one 9/11 per year, which is terribly unlikely. So that's a little over $2 million per life saved, as an extreme low-end estimate. It's probably more like $200 million per life saved.

Now, how does that compare to other means of saving lives? Are we getting a good deal?

First, the number of lives involved. Less than five thousand Americans have been killed by terrorists in the last thirty years. Compare that to annual death rates of tobacco use, which is something like 400,000. Or 43,000 for traffic collisions. 85,000 for alcohol. 17,000 for drug abuse. Or 41,000 suicides. Looking at the leading causes of death in the United States, it's clear that terrorism shouldn't even enter your mind as possibly leading to your death.

So we have our numbers of deaths. Let's start with those who die of alcohol and drug abuse. Presumably at least some of those lives would be saved if the drug abusers got into rehab. So say we're extremely generous, and pay for six months of rehab for every drug and alcohol abuser. How much would that cost? Roughly $28,000 per month gives us $168,000 per life saved. Even supposing only a tenth of those given treatment actually come out the other side more likely to survive, that's still $1.7 million per life saved. Better deal than the TSA. Suicides could probably benefit from similar treatments being available.

Or how about motor vehicle accidents? Something like 16,000 people die each year from drunk driving alone. I could come up with all sorts of complex solutions, but suppose we just gave out free taxi rides to drunk people? That's about as dumb and expensive as it gets, right? Say it's $100 a ride, a million rides a week, so five billion dollars a year. Even if we only prevent 20% of drunk driving deaths, that's $1.6 million per life saved. Still a better deal than the TSA on it's best imaginable day.

And how about tobacco-related deaths? Let's again throw money at the problem: free vape for everyone! According to one very biased source, a vaper might spend $600/year. Say it's $1,000. Suppose everyone who smokes took advantage of that, and that we actually created even more users since it's now free. Call it $70 billion a year. And suppose we prevent just 10% of smoking deaths in a year, or 40,000. The inefficient, expensive, barely effective plan I've just described is still more cost-effective at saving lives than the TSA!

Let's have some fun and look at colorectal cancer! 41,000 lives a year could be saved by appropriate screening. Like everything in the American medical system, pricing is completely opaque. Uninsured cash price for a colonoscopy is, at the high end, $5,400. It'll probably be less. Suppose we pay out $6,500 per procedure, and let the patient keep the difference as a cash incentive. Say that everyone over the age of 50 has a colonoscopy every five years, twice as often as recommended. So that's twelve million colonoscopies a year. We spend $78 billion dollars a year to save 41,000 lives. That's still cheaper than the TSA!

On a dollars-per-life basis, we could do much more good with the TSA's budget, even with incredibly stupid ideas like these.