Immigration chain addendum

Our chief idiot is floating that he will strike down birthright citizenship with an executive order. It's probably all hot air, and would be struck down in the courts, but it definitely stokes fears of immigrants. One of those fears is that parents will come to the USA just to have their kids and make their own path to citizenship easier.

While this is sort of maybe technically true, let's look at the details. As I've outlined in the past, the "links" in "chain immigrations" are prohibitively long. To recap, a US Citizen of age 21 or higher can petition for their parent to be given a green card. Using the anchor baby as the basis, this baby will need to live for 21 years, then petition, then wait at least 7 years, but usually 20+ (especially for those from countries that certain people tend to worry about). 40-45 years after giving birth, a mother can be reunited with their child in the USA. This is an exceedingly long play: the mom has a decent chance of not living that long, and the people worried about this abuse are even less likely to be alive by the time it happens.

Another concern is they are going to leach off American benefits. If the kid stays in the country, then yes, I suppose they will … but they will grow up in America like any other kid, will probably get a job like any other kid, etc. In other words, they'll be like any other kid and should get the same exact support. It's economically sound, and morally right. If they don't stay in the country as a minor, then they won't get any of the benefits and problem solved. 

Working on the immigration chain

"Chain immigration" is now a common topic of discussion. How does it work?

Claim: Well, anyone in America can just bring their random relative over.
Fact: The family-based visa program is limited to nuclear family: parents, children, sibling, spouse.
Fact: The family-based visa program is limited to sponsors who are already US Citizens.
Fact: The family-based visa program allows a US Citizen to _petition_ for a visa for a close family member. The applicant must be background checked, interviewed, etc. There is no guarantee they will be granted a visa.
Fact: Once approved, the family member receives a green card and can stay permanently in the USA.
https://travel.state.gov/…/family-based-immigrant-visas.htm…

Claim: Ok fine, it's limited, but then people arrive and just repeat the process.
Fact: Remember, sponsors must be US Citizens. It takes a minimum of 3 years to receive citizenship after arriving and getting a green card.
Fact: Family-based visas are limited in number (currently about 200k are given out in a year). This means there are backed up waiting lines, often in the 5-10 year range (and more for some countries). Your married child or sibling from Mexico or the Philippines would have been waiting for 20+ years to be approved today.
Fact: The only group that is approved faster (about 2 years) are unmarried minor children.
https://immigrationroad.com/vis…/visa-bulletin-by-month.php…

Claim: But the chain! They will just work around the law by getting their parent, who will get their parent, who will get their other child, who will get their child, and now the cousin is here. Also I can't do math.
Fact: I will do the math for you. It will take a US Citizen, who is at least 21, at least 7 years to get their parent in the country. That parent will then require another 3-5 years to get a citizenship. So, in 10-12 years, they can sponsor their parent and there's no guarantee it will be approved. By the time it might be approved and your grandparent immigrates and becomes a citizen, you are at least 41 and your grandparent probably isn't alive anymore. But if they are, they can try to repeat the process twice more, and 20-25 years later your cousin is here. You are at least 61 by this time. Also, these are the absolute best-case times.

Claim: Oh, I got it, instead of parent to parent to child to child, we can do parent to sibling to child. That eliminates one step.
Fact: Yay, a critical think! Now the best case is your cousin can be here in 30-35 years.

Claim: Actually it's 25 cuz the cousin will be here after 5 years, you counted 5 more years for them to get a citizenship but they're already here and have done all the crimes so nana nana nana.
Fact: You got me there. Everything I've said must be invalidated then. Unless I remind you that 7 years for the wait is the best-case, and good luck with that. For Mexico, each step will take 25 years, so by the time your cousin gets here, you'll be 100 years old and you won't even remember them.

So how does it work? It doesn't. Not in any real practical terms.

How does society advance

We were once cavemen. We would spend our days hunting mammoth and gathering berries or whatever, and that was that. We banded together with others we could trust so that mammoth hunts would succeed, but our daily life was still this loop: find it, get it, eat it. Repeat. Occasionally fight off an intruding sabre-tooth tiger or bear or giant beaver. There was little time for anything else.

Eventually civilizations rose. Their hallmarks were governing bodies and the ability to grow and store food. Without having to worry about where calories would come from each day, people could spend their time doing other things: organizing group efforts to build structures, developing science to help with that, learning how to build tools and ride camels, etc. Society became more functional because people had more time for activities beyond basic subsistence.

Even today, the next great invention or startup or idea will materialize because someone has the time and freedom for it. We can best elevate society by removing or lowering subsistence barriers as much as possible.

Rampant deregulation flies in the face of this. It allows some who are innately aware of certain things to thrive. Suppose we don't have a guarantee of clean water. Those of us who know how to test it will take some time to do that and then go about our lives, and the rest of us will have lots of gastrointestinal or death issues. All of us will waste a tremendous amount of time either testing the water or suffering from its effects and be unable to do actually productive things.

A significant measure of society is its cumulative productivity. Productivity requires food, health, education, and a low enough baseline stress level to function. Said stress level is often related to the availability of food, health, and education.

Maybe we need to create a food voucher program for everyone, not just the needy. Remove the conditional administration overhead and just give everyone a FedFood card that reloads every month, no questions asked. If you want to buy something fancy and have the money for it? Cool, do that. Isn't that some crazy huge tax though? Math time:

1. The USDA estimates the basic cost to feed an adult is about $50-60 per week.
2. Approximately 12-14% of USA households are food insecure

The shortage in dollars is, therefore, at most $50-60 per week * 12-14% = $6-8 per week, or $300-400 per person. Most food insecure households have some money for food, so the actual number is significantly lower. This is how much tax redistribution, per person, it would take to make sure everyone has enough money to eat. The average tax per person should be about the $50-60 per week. For those of us who already can afford all our food, a portion of our expenses just makes a trip to FedFood via these taxes. That money is even tax-advantaged (in my hypothetical system), meaning many of us would come out net better - I get $3000 a year less in net pay, but don't pay taxes on it either. If I get it back as a tax-free spendable credit, I've come out ahead even when adding the several hundred dollars overhead to cover those in need.

We can make a similar argument for healthcare; this is a well-discussed area.

Betsy DeVos is currently trying to go the opposite route in education to make people like her richer. Instead of undermining one of our basic societal needs, she should just go stick her $1B net worth in stock market and buy a new yacht every year. We can instead spend a fraction of the $600B annual defense budget on federal teacher resource supplements: no more buying crayons for their classrooms out of their own paychecks. Or just give them bigger paychecks.

Clearly these are not carefully thought through end-to-end, but maybe there's something there. As a society, we should find the issues that negatively affect most of our members' ability to lead productive lives, and address them. Repeat.

Parallel Universes

This is not about physics. Or time. It's about lies.

When a person lies, they create a parallel universe. In this universe, some fact is different.
The person now has to take care to live in both universes: the real one with the people they haven't lied to, and the alternate one with the rest.

But don't alternate realities collapse when the main character meets themselves? In the same way Marty McFly can't let other Marty McFly see him, interactions between those in the real and alternate universes are fraught with peril. If they talk to each other and discuss their versions of the same fact, the universes become incompatible.

Every time the person lies, they create more and more parallel universes and have to maintain each one. Juggling them is hard work, eventually some will drop. The only universe that stays stable is the one with all truth. In the long run, reality wins out. Lies requires additional energy and become unsustainable.

Unfortunately "the long run" can be quite long on a human lifespan scale.

The Paul Manafort phenomenon

Paul Manafort was in deep trouble, under house arrest and investigation. Then, he decided to try and tamper with witnesses and is in even more trouble now. Why would he do such a thing? He must be very savvy to have made it as far as he has.

I think career criminals often have a risk-taking streak. They would almost have to, to justify all the risk they take to get their reward. This personality trait is just showing up, still: he sees reward in colluding testimony and has the constitution to take it.

Shouldn't he have gotten caught already? Well, no. Criminals come in all levels of capability and all different tolerances for risk. Those who succeed at the best chosen risk-reward scenarios tend to do the best. But, every act carries some odds of failure. A criminal can only fail so many times before they spend their life in prison.

There will be some criminals who aren't particularly good at what they do, but the odds happen to fall in their favor. Just because they've made it far doesn't mean they are great at what they do. 

The Warren Buffett phenomenon

I feel like I've written about this before, but a search of my blog shows no hits for any of the words in the title. No one's reading anyways, so it's cool if I repeat myself.

Is Warren Buffett the greatest investor ever? Many would say so. He certainly has a very long, proven, track record of successful investing. His Berk/A stock has risen from about $7000 a shared in 1990 to $67000 a share in 20 years ago to $287,000 today, easily outpacing the S&P 500 index has gone from $350 to $1000 to $2800 in the same time period.

Does this mean he has the most insight into investments? Not necessarily.
Investment is about taking good risk-reward positions time and time again. It's about assessing the risks and potentials, then acting. However, no one knows the future for sure so some of it is luck in how things unfold. Buffett is almost definitely a very good investor, but he may not be the best. Rather, he's the one for whom the future unfolded just right. And there is always such a person.

Warren Buffett is an illustration of how someone can rise to the top but may not be the best at their craft. This is critically important to remember when luck is involved in success. Being good improves your odds, and someone has to get the best result in a competition. Those two may not be the same.

DACA notes

Our President today claimed that all these flows of people are trying to take advantage of DACA. Who knows, maybe he's April Foolin'. Realistically he's just straight bullshitting.

The eligibility considerations alone are not well understood by the broad public. A DACA recipient must:

1. Be born on or after June 16, 1981
2. Have resided in the USA since June 15, 2007
3. Have been physically present in the USA on June 15, 2012
4. Be in school or already in possession of a high-school diploma or GED
5. Not have criminal history beyond 1 or 2 misdemeanors
6. Be at least 15 years old.
7. Came to the USA before turning 16.

There are objectively true things we can conclude from this.

The most important is that potential DACA status applies to a fixed population. While not everyone who is potentially eligible has applied, no one coming into the USA after DACA went into effect is eligible. Any narrative about family bringing in young children to take advantage of it is completely false.

Another serious point is that it can only apply to people who reasonably came to the USA without understanding the illegality of it, or at least were not in a position to make a judgment about it. While they came here illegally, it's highly debatable if they themselves actually committed a crime (I'd tend to say no). They lack(ed) legal status, but are not criminals.

They are required to complete, or be in the process of completing, at least a high school education. If they drop out, their DACA status is lost and they become eligible for deportation. This is a higher standard than we hold our own citizens to.

They can't commit a felony or serious misdemeanor, or be a repeat minor offender. If they do, status is lost and deportation can follow. This is a higher standard than we hold our own citizens to.

If DACA stops taking new applications today, anyone eligible who was less than about 5 years old will not be able to apply (since they would not yet be 15 today). These people probably have no memory of their home country, there's a decent chance they don't speak their home language to any level of proficiency, and probably grew up as "American" as anyone else here. Current rumblings say DACA applications will cease early next year (this is the '6 months' Trump is talking about), leaving this most vulnerable population the most affected.

The convoluted art of the timeshare deal

We just returned from our third trip to Maui. The impetus for the trip was a timeshare presentation offer we couldn't refuse: in exchange for a 90 minute pitch, we'd get lodging at the Hyatt Residence Club on Ka'anapali for $1295 plus taxes and a $100 spending credit while we're there. The taxes were about $180 dollars and we would no doubt charge at least a few things to the room, so the net was 5 nights in a fantastic 2-bedroom condo with ocean view, right on the beach, for just under $1400. We split the cost with my sister-in-law's family, bringing our net to about $140 a night per couple, per night, in premium accommodations.

The presentation was scheduled for the 3rd full day of our stay, giving us enough time to enjoy everything and be in a fully receptive mood. Our agent took 10-15 minutes to get to know us, asked about our travel and spending habits. He did some math to show us we'd spend about $500k in a lifetime of vacationing. This, of course, is meant to make the ensuing offer seem cheap by comparison (as he writes $500k in one column and $2600 per year in the other - we'll get to the actual math later). He lays out that we have 4 options: use our purchased week, trade it in for 2600 points towards another Hyatt Residence Club location, 3000 Interval International points, or 90,000 Hyatt hotel points. These are, of course, all different opaque points in different systems, with different nebulous values.

He then takes us around the property to show us professional shots of the Residence Club properties and declares that Interval International has 3000 hotels in 80 countries (which is supposed to make you think "I can go anywhere I want!"). We are presented with another scale involving Travel Destination Index (a measure of the desirability of a location at a time) and bedrooms pointing out points per week, and he highlights that by trading in our week in Maui, we could get 4 or 5 weeks in some other locations. He also tells us just how high the retail price is these locations we'd get access too. For example, our Maui condo would run $8800 a week after taxes (but could be ours for just $2600 a year!). There were many other leading questions about "wouldn't it be nice to just show up and not have a bill?" and similar. Sure, sounds great, but how much do I pay ahead of time for it? We eventually tour the grounds of the timeshare property and gaze over the ocean from the 12th story. He tells us how much courage the first owners had to buy before the property was even finished ... and did we have the same courage to become owners too?

What's the deal?
In the end we said no, partly because there's simply no way to assess such a convoluted asset in an hour, and mostly we weren't interested to begin with. In the interest of keeping an open mind (as we had been asked to do), I wanted to work through what the offered asset is actually worth, and explore the value of the offer. The terms are:

1. Buyer makes a down payment of $66-76k, depending on the chosen floor (2 to 12).
2. Buyer pays a yearly HOA payment starting at $2600 (assume it rises over time)
3. Buyer may, in each calendar year:
1. use the property for any 1 week, excluding Christmas and New Years (which cost unspecified extra)
2. exchange use of property for 2600 Hyatt Residence Club points
3. exchange use of property for 3000 Interval International points
4. exchange use of property for 90,000 Hyatt Global (hotel) points

The true cost to the owner is two-fold: there is the opportunity cost of the down payment, and the ongoing cost of the HOA fee. Qualitative opportunity cost, such as being able to afford a primary home, doesn't belong in this discussion; I will assume anyone considering this has $70k free money to put down. The opportunity cost can be estimated quantitatively by looking at lost investment opportunity. Conservative estimates would say about $2800 per year, optimistic estimates could reasonably be as high as $5600. I'd tend to bracket the range as about $3500-5000 per year. The HOA fee will rise as costs rise. Our agent claimed accommodation costs were rising by 10-15% per year in the area, which is believable when comparing our costs to stay at the Hyatt Regency next door in 2012 to their published rates today. There may also be a one-time large inflation as travelers have shifted away from Mexico due to Zika virus and cartel-related violence. At any rate, let's assume 10% long-term. In early years, the actual cost is about $6000-7500, and will grow to $10-12k in a decade.
The true benefit can be quantified as the market value of whichever option in [3] the owner takes that year. Each sub-option deserves its own analysis. In other words, how much cash would I need to spend to get the same thing?

Value of property use
I can't currently find a published price to rent a week at the Maui property, so the only number I have is the claimed $8800 after taxes. However, we must also compare similar nearby options (unless the Hyatt is uniquely special to someone). There are several other condo buildings on Ka'anapali, along with numerous hotels. A year ago we rented a 2-bed, 2-bath vrbo set a quarter mile back from the beach, with sweeping ocean views and lanai. We watched whales every morning over breakfast, had a pool, hot tub, kitchen. We paid just under $3000 for 7 nights. The actual place was a little less posh, but still nice, and we didn't have room service or a bar on premise ... so we took a trip to Costco when we landed instead.
For comparison, the nearby ocean-front Ka'anapali Alii runs between $500-600 a night for a 2-bedroom condo. The building is clearly older and the units are a bit outdated, but we have to consider the relative value.
A comparably appointed Westin or Hyatt hotel on the beach, with ocean views runs about $400-500 per night, and this cost would be doubled to get to two bedrooms. This does not appear to be a better value. Opting out of the ocean views would save about $100 per night.

Value of Hyatt Residence Points
The additional consideration here is the limited set of locations. This exchange only works if there's one in a desired destination. There are 15 other Residence Clubs, and they are heavily concentrated on ski (Tahoe, Colorado) and beach locations (with 4 in the Florida Keys), and then a few more esoteric spots like San Antonio and Sedona. There is no indication on their website of how many points any of these could be booked for, that information seems limited to owners. Our one data point is that we rented a 2-bedroom at the Hyatt Sedona Residence Club, in August 2013, for $282 a night. That is clearly far under the market value of the Ka'anapali property. Searching "Hyatt Residence Club" on hotels.com offers some glimpses, for example the Hyatt Beach House in Key West is indicated "from $319 a night", but no concrete bookable offers popped up in my quick searches. At best I can conclude that comparable quality properties can be had for that range, so unless the points can be translated to longer stays, it's a clearly lesser value than just using the home property.

Value of Interval International Points
There are 3000 locations in 80 countries. While talking, Marisa thumbed through the booklet and noted that many of the locations seemed to be in slightly obscure spots. For example, there were half a dozen options in Hungary, but none in Budapest. A quick perusal suggests the locations tend towards the more remote, though this is not conclusive. While the website allows a search of locations, you need to be a member to see any cost details. This lack of access makes this entirely opaque, so we can't really evaluate the value.

Value of Hyatt Global Points
90,000 sounds like a lot, but nights seem to start at 8000 per, and a mid-range Hyatt hotel is about 12,000 (per examples shown to us). Nights start as low as 5000, but I expect these are few and far between. Even if a 12,000 point mid-range hotel is $300 a night, that's still only a $2000-2500 benefit in exchange for the original property.

Are people buying?
Yes. Assuming our agent was truthful, about 50% of the timeshare has been sold. There are 52 weeks times 134 units for a total of about 7000 shares. The property has been open for about 3 years so they are selling about 3 shares a day (3500 divided by ~1100 days). Assuming every buyer is getting a pitch in person, about 6% of the claimed 50 daily sessions are converted into sales. The bulk of the sessions are set up in exchange for freebies like sunset cruises or golf, my estimate is that 5-10 of the daily sessions are for people who got a deal similar to ours.

Is it worth it?
To me it appears this is not generally worth it. For someone who has 70k and nothing else to do with it, who loves the nicest accommodations and Ka'anapali, it's not an inherently terrible deal. Using $6-8k as the effective cost, comparing it to $4500-5000 for a less nice Ka'anapali Alii, or $7000+ for comparable hotel rooms, the numbers aren't way off. Assuming the retail prices grow at 10% for the area and the HOA grows at the same rate, locking in the bulk of the investment as a one-time payment can be a benefit. It does not appear that the exchanges are great options, so the value of the asset is primarily the one location. Anyone who prioritizes flexibility and variety in vacations is probably better off saying no. Still, it's amazing to me that people are committing $70k based on an hour of cherry-picked information. People ask questions, but not all the questions. However, having been able to ask a few questions might make them feel comfortable? It takes way longer to assess the underlying asset. I just spent several hours laying it out and trying to dig up representative numbers and wasn't able to answer questions to my own satisfaction.

What about buyer's remorse?
The presentation is short enough (and we weren't interested enough) that we never even asked about selling. Apparently it's possible to sell back to Hyatt at the current market rate minus 6% commission. Those rates are not made public, though I suspect it'd be easy enough to find out what the current asking price is. I suppose that's better than lots of other cases where people are desperate enough to get out that they just abandon their asset and lose everything, but that's a pretty low bar to beat.

Save our children. The more the merrier.

The Parkland shooting has created a momentum in the conversation around gun and school safety reform that hadn't really existed after previous shootings. The students held rallies, addressed lawmakers, created social media personalities, and lots of people followed the cause. 17 of their classmates and teachers had just died, it was time to do something.

The Florida legislature is in the process of passing some reform: a combination of raising the minimum age to purchase a gun to 21, 3-day waiting periods for all rifles, and $400M for mental health and school resources. The resources can include a variety of defensive measures like bulletproof glass, metal detectors, etc. All this sounds pretty good from a reform perspective, right? There was one other nugget in there, and it's a point that has also gained a lot of traction: putting guns in schools. In the Florida legislation, it would mean allowing "trained" teachers to carry guns unless their school opted into a marshal program. I hadn't seen the details of the latter, but presumably it's essentially putting a full-time, uniformed police officer in the school. Around 30% of the approximately 100,000 schools in America have such a School Resource Officer.

Is this a good idea? There are numerous lines of reasoning. One says that our children's safety is paramount, and adding a trained, qualified, armed person to their safety net is a win. Another says the SROs can lead to some unintended consequences, for example unruly children might end up with criminal charges instead of detention. There's particular concern the latter will apply for minority students. I have an entirely different concern: finite resource optimization.

We have a limited amount of funds to do all the things we'd like to do. This is true of individual budgets, state budgets, school budgets, all budgets (even federal, though they have a bit more leeway to take on debt). The goal of the SRO is noble: to save children's lives. No one contends this is a desirable outcome. However, we need to consider the cost-benefit not because we are putting a dollar amount on a life, but because we want our finite resources to protect the most lives. To compare one approach to another, we have to quantify the cost and the benefit: how many lives are saved for how many dollars, and can we deploy those dollars in a more effective way.

An SRO is a police officer. Salaries will vary, but the cost to support an officer should be at least $75k per year between salary, training, equipment, etc, amounting to around $7-10B per year if deployed to every school in the nation. Let us also suppose, optimistically, the tactic is completely effective: all school shootings are eliminated, no child dies from gun violence in a school, ever again. Let us even assume that school shooters would not be displaced to other locations, ie: that they won't become mall or park or clock tower shooters instead. We would save an average of 20 lives a year, with a high mark of 42 in 2012. We'd spend about $350-500M per saved life. Again, this is if everything goes perfectly. A similar number injuries would also, optimistically, be prevented.

There are an abundance of ways we can save lives, including children's lives, for far less. Some are investments: teach them more about the dangers of addictions, alcohol, drugs, gangs, HIV. Some are more acute: Spend the money on cancer research or leading-edge treatment to offset some of the two thousand childhood cancer deaths each year, or set up a free no-questions-asked cab fund for teenagers to use when they have been drinking. Hell, make it free for everyone. Create a monetary incentive for blood donations, or run ad campaigns for people to become organ donors. Provide housing for the homeless. It doesn't take much imagination to do better than several hundred million dollars per life.

Guns, guns, and different guns

On a very basic level, all guns are the same. I'll let Homer Simpson explain:



However, they diverge rapidly from there. In all the discussions following various mass shootings, the anti-gun (or gun control, if you will) crowd has jumped straight to policy with very little understanding of how guns are different from each other. There are many calls to "ban all those automatic and semi-automatic guns". Of course this warrants immediate contempt from 2nd Amendment advocates (or pro-gun, if you will) because it reveals a clear unfamiliarity with the world of firearms. "Semi-automatic" seems to be a loose term for "really powerful" and I'm not sure what automatic is even supposed to mean in this context ... but maybe "military powerful"? At any rate, these terms have concrete meanings and they matter to those who understand them. In the interest of dialog, it's important that both sides use proper terminology.

Automatic means the user pulls the trigger once, and bullets keep flying out until s/he lets go. These are what we usually see in our action movies, and include the M-16 and M-4 most commonly used by our military in war zones. They are able to fire hundreds of rounds per minute, in principle an M-16 could empty its 30-round clip in just a few seconds. Larger belt-fed models like the M-60 can sustain that rate much longer. Not all automatics are big or powerful; Uzis are a prime example of a small automatic fire weapon. These guns are largely illegal in the civilian population due to the 1986 Firearm Owners' Protection Act.

Semi-automatics require the user to pull the trigger once per bullet. That is the only fundamental differentiation from the group above. On a per-bullet basis, these guns can be just as powerful as those above. For example, high-powered hunting rifles are available in semi-automatic. The AR-15 is the predecessor to the M-16 and M-4 above and shares their firepower. However, many are much smaller: the Glocks used by many police forces and civilians, or even this tiny Beretta.

There are also guns that require additional input from the user to load the next round. Old single-action revolvers required the user to pull the hammer back to turn the cylinder. Bolt-action (common for hunting rifles) requires the user to explicitly eject the previous cartridge and chamber the next round.

What actually damages a human? It's the energy transferred by every bullet that hits them. This energy can be quantified: it is the "muzzle energy". For example, an M-16 uses a standard NATO round that is rated around 1800J, while a 9mm Glock might use a .380 ACP round rated around 300J. Even big-boy handgun ammo like the .45 ACP and 10mm auto top out around 800-1000J. As this trauma radiologist points out in a well-circulated piece in The Atlantic, the impacts from these are worlds apart.

In addition to the damage caused per bullet, there's the matter of how fast accurate shots can be fired. The AR-15 looks similar to military guns precisely because it was the original, designed to do the same things. It was built with the ability to take a large magazine, to be aimed easily over long distances, and to recoil minimally so the process can be repeated as fast as possible. While not a fully automatic, an AR-15 still has an effective fire rate around 1 shot per second. Since each bullet causes so much damage, the 20 or 30 in the magazine can be targeted at that many separate targets. Because the weapon is so much more accurate and easy to aim, a skilled user can reasonably expect to deliver a lethal shot from 500 yards, vs maybe 50 yards for the Glocks used by police. Given a crowd, a skilled operator could probably kill a dozen people in under 30 seconds. This combination of power, range, and accuracy makes it hard for a crowd to get to safety. It allows a shooter to set up in Mandalay Bay and rain death onto a concert hundreds of yards away.

Interestingly, recoil is a big enough issue in handguns that the FBI requested lower-speed ammunition for their Glocks outfitted with .40 S&W ammo specifically to improve their weapons handling. Because properly used rifles eliminate recoil to such a huge degree, they can retain their rapid fire rate and high power.

It's easy to see why mass shooters are drawn to "assault rifles". Interestingly, the AR-15 (or non-Colt-brand variants thereof) used in many shootings does not qualify under the US Army's definition since it lacks selective fire capability (it's only semi-automatic). The definition doesn't matter. The function does. It fires a highly lethal NATO round that was designed to penetrate 1/8" steel armor and steel helmets at 500 yards. There's no reasonable need for that level of firepower outside of a war zone. Upping the ante, the highest firepower available is the 18,000J of the .50 BMG round, used by the readily available Barrett M82, with effective range over a mile (and confirmed longest deadly shot at 2.2 miles). Thankfully it makes a poor "assault" weapon because it weighs 30lbs and rounds cost about $4 apiece. But, it would be highly effective and lethal from a perch.

Attempts at legislation can be obviously tricky. Attempts to outlaw ammunition above a certain muzzle energy will impinge on big-game hunters. Shooting a moose, elk, bear, etc, requires serious firepower. Limiting specific weapons or ammunitions will just cause manufacturers to create slight variants of existing models. The trick is to find the right feature, or combination of features that eliminates weapons with mass offensive capability. I think the right ideal model is some combination of limiting muzzle energy in combination with semi-automatic weapons. For example, suppose you can't buy a semi-automatic weapon that fires rounds over 1000J. This ends up being unenforceable through the combination of a custom "weak" round for the semi-auto weapon and the creation (or existence) of a bolt-action gun taking the same format. There's some precedent already via the definition of all illegal destructive devices, which does include any ammunition with great than 1/2" caliber ... but the law mostly focuses on explosive devices, rockets, etc. There's a bit of a "you know it when you see it", but that can't translate to law. Maybe a solution is to eliminate all semi-automatic rifles in an attempt to limit the rate at which damage can be done?

The 1994 Federal Assault Weapons Ban was an attempt to classify what an "assault weapon" is, but its guidelines seem to have a lot of loopholes or misses. For example, rifles had to meet at least two of:

• Folding or telescoping stock
• Pistol grip
• Bayonet mount
• Suppressor or threaded barrel to accept one
• Grenade launcher

It's pretty easy to imagine an AR-15 variant without suppressor threads that doesn't meet the above definition. This illustrates that even with the appetite to make changes, and even if an AR-15 variant is explicitly disallowed, it would simply require the manufacturer to make a new model with a minor cosmetic change that would not be covered under the legislation. Perhaps the approach could be that new models must be opted into legal status, and existing models can be placed on disallowed lists?

It's also important to remember that mass shootings make up just a few percent of all murders. Handguns are responsible for the majority, but they rarely make headlines because they usually just kill one or two people at a time. The goal of limiting "assault rifles" is just to limit the carnage in special cases with deranged shooters. Any new laws need to address issues while not unduly burdening citizens. Proposals like arming teachers (and training them, and spending taxpayer money on more guns, ammo, etc) definitely seems to fail that. But anything we can try for "cheap" that might improve our outcomes should be tried, measured, evaluated. And then we build from there.

That big, beautiful, idiotic wall

Trump wants $18B over the next decade to build some new border wall and reinforce old walls. CBP wants funding for additional agents and equipment, bringing the total to $33B. ICE wants to add 10k more agents as well, which has to cost at least $1-2B per year. Grand total: about $50B in the next 10 years.

After all that money is spent, there will still be half a US-Mexico border sitting naked. We'd need to then spend the same money again, two or three times over. While a few areas don't need a wall, most does.

There is a lot of debate about how much illegal immigrants cost the country. High estimates touted by Trump and Fox personalities claim $113B per year, but this has been rated mostly false by Politifact. I don't entirely understand how these numbers are computed, but the CBO's assessment is that "in aggregate and over the long term, tax revenues of all types generated by immigrants—both legal and unauthorized—exceed the cost of the services they use". This may include the general economic activity generated, though it sounds like a direct comparison of just net tax dollar flow. Either way, if the reality is that illegal immigrants are revenue-neutral, then spending money to keep them out is bad financial business. Consider also that illegals interact with legal businesses and generate revenue for those as well (they eat, they live places, they need clothes, tools, cars, ...).

The effort to keep them out becomes a moral question, but even there it can't be assessed in a vacuum. We must compare what else we could do with the money. $5B could cover ongoing health care for 500k people. Or a yearly $150 tax break for the bottom 10%. Or a $1500 raise for every teacher. Or installing 1500 new wind turbines a year that will offset fossil fuel energy used by about a million people. Or a full-ride state university scholarship for 100k students per year. It could be an investment in our own people. And consider that even a perfect wall would only keep out half the illegal immigration (the other half enter legally, then overstay visas).

The common line is that being anti-wall is the same as wanting weak borders. No one is saying we should promote illegal immigration, but we have to consider the cost-benefit or the morality here. In either case, it seems more prudent to spend the money on something else. Anything else.