lewis500

The number of jobs involved in this scam is trivial either way. The point of solar power manufacturing is not to make up jobs for people to do; it is to convert our energy to a source that doesn't kill people. If our planet is really at a slight risk of danger, and if solar panels somewhat ameliorate that, then it is stupid to prize a few thousand jobs. Last month the economy created a net 233k jobs---in one month. The economy is $19 trillion and we are waging a PR war over $1B vs $2.5B.

Federal reserve and fiscal policy are 1e9 times as important to job creation as this tariff. Would it be bad for the economy if someone invented a machine to make solar panels cheaply without any labor at all? No. That would be good because we could have lots of solar panels and save the planet.

Important to remember these tariffs were not conceived of as a job creation scheme. They were created to produce a windfall for investors who bought the bonds of bankrupt US solar companies. https://www.bloomberg.com/news/articles/2018-01-23/jpmorgan-...

I think that, for the White House, the tariffs have an added benefit of quashing an industry that more and more competes with coal. For reasons of graft, an obnoxious "heartland" brand and some weird fetishization, this administration is deeply obsessed with coal---one of the worst industries in the country to work for, which kills countless people via pollution.

You will have Social Security. You just might receive something closer to the current level of social security benefits rather than a more generous package. Social Security benefits are indexed to median wages, which rise over time. If they just don't increase benefits so quickly system will be fine. Likewise, a hike in taxes which would still leave us one of the lower-taxed industrialized countries would do the trick.

A funny thing is that we have both fear of automation and fear of insolvent social security. If the economy gets more automated it will grow quickly, and it will be easy to tax the additional income to pay for social security---although we may have to use a value-added tax or a higher corporate tax rather than payroll taxes.

I wonder how much empowered workers would bargain for wages vs other things they might want. For example, in the US, we public employees do bargain for higher wages, but we also go pretty hard for pensions, job security and lower health co-pays. Increasingly we are also bargaining for childcare help due to the insane increases in the cost of childcare (this is a genuine mystery to me: in the 80's early 90's my broke parents could easily afford to put me in full-time, quality childcare from like 3 months onward). In European countries with strong bargaining, workers demand more vacation time.

In short, I don't think it's right to imagine wages going up enormously with more bargaining power; I think it's more accurate to imagine benefits and terms of employment (job security, vacation, autonomy) going up. Not a normative point but more of a positive one.

I believe this happens because of how the union works. People who care a lot about some issue (e.g., parents caring about childcare) will push hard on the union to bargain for their particular benefit. Wage increases, by contrast, are spread over the whole union and thus expensive in total but not huge for any individual, so there is a sort of collective action inertia even within the union. A small wage increase would cost as much as a huge increase in childcare help, so the parents will be able to mobilize more easily. Personally, this arrangement me benefits me a lot as I care more about those benefits than salary.

Only as I get older do I realize how incredibly bad elements of law enforcement in the US can be, and legally so. Few cops and agents are bad, but it is a system with little filtering, personally or institutionally, of the bad elements. Civil asset forfeiture is overwhelmingly unpopular, but when states try to reform it the law enforcement groups rally around it mightily. Essentially they want the right to steal money because they need money; nobody really makes a sincere argument the situation is any different than that. In Alabama they recently tried to reform it. In an op ed by the leaders of district attorneys and sheriffs, the authors wrote “What incentive would local police and sheriffs have to invest manpower, resources and time in these operations if they don't receive proceeds to cover their costs?” I dunno: it seems like if they are professional cops they should just solve crimes and such that seem important; by “investing manpower” they are not investing like a firm does but pulling resources from elsewhere to spend trying to take things from oblivious people. Also in alabama, sheriffs are allowed to take extra money from the food programs at the county jail, and they buy themselves vacation homes this way. https://www.cbsnews.com/news/alabama-sheriff-legally-pockete...

When I was in grad school I made these to explain concepts from traffic theory:

http://setosa.io/blog/2014/09/02/gridlock/ http://setosa.io/bus/

These were major efforts and I don't recommend building out anything similar while pursuing a master's. But control theory does lend itself to simulation well, so I would use jupyter notebooks to present little annotated simulations with code. Print out plots of the state space at intervals for example. I also highly recommend the graphing calculator desmos.com.

Bruce burrowed into a floating piece of the asteroid and now has a sick man cave. Remember he is the best oil driller in the world.

The diamonds are cooled by ice water

Armageddon 2: Starmageddon. Bruce Willis has to land on the surface of a star, drill down into it and release a hydrogen bomb. The fusion reaction will cause the star to blow up even though it’s already full of fusion. He’ll wear a suit made out of diamonds and have a spaceship made of diamonds, and an Aerosmith rendition of twinkle twinkle little star will play as the star explodes and diamonds rain from the sky. Now seeking story credit for this script.

Actually military figures are fine for this expenditure because they are very large per unit, so easy to conceptualize in small integers. It doesn’t matter if you can think about how complicated they are to build or their deterrent role or whatever. Point is you can imagine an aircraft carrier. This is about as expensive as 1.5 aircraft carriers or three bay bridge replacements.

About 1.5 aircraft carriers or 3 bay bridge eastern span replacements. I have various costs in my head to conceptualize how much things cost in terms of what we could’ve bought. The last aircraft carrier the US bought cost about $13 billion. The bay bridge eastern span cost 6.4 billion.

For smaller costs a good one is submarines: about $2.4 billion.

I did the math once and the war in Iraq directly cost about the replacement value of the nyc subway but can’t recall the figures well.

Interesting to see this back on the front page after three years. Still remember us sitting in our living room drawing this on paper and arguing about the right approaches.

Maybe one day vicapow and I will make a triumphant return to the explorables space, but life has a way of getting in the way as you get older.

better public transportation would only alleviate the problem if it goes to places where additional supply gets built. otherwise it will just drive up prices in the places to which it adds accessibility. imagine a teleporter in stockton that goes anywhere in the bay. unless stockton allows additional housing near the teleporter, it will just become even more expensive than san francisco.

"if threats of a trade war with China or other countries turn into reality, this will tend to increase inflation, driving up prices both directly through new tariffs on imported goods and indirectly by encouraging less efficient production."

I wouldn't call this "inflation" in the sense economists mean it. This could cause a one-time rise in prices, but it would not necessarily mean permanent change in the rate prices go up each year.

Maybe some of yall would like this federal reserve simulator I made last year: http://lewis500.github.io/macro/ It is one of my least popular explorable explanation but also one of my favorites.

Bless you

I want to know how this does with traffic. Pretty remarkable.

I am pretty mystified why tesla is even running autopilot on their cars. They already have more demand than they can handle, and I don't think anyone buys a tesla because of the autopilot. They are opening themselves up to huge court claims at the same time they are low on money. Just keep running tests and maybe do a pilot for a few years. Isn't Waymo sort of in the lead right now? They aren't running their software on millions of vehicles but seem to be progressing okay.

It seems all these companies are in a big rush and being slapdash. Maybe it's a disconnect between the engineers and the execs/shareholders. It doesn't even seem profit-driven...almost fear driven. ("we don't wanna be left behind.")

This is how Comcast works too: introductory offer and then the price goes up. So about every two years I call them acting like an over sharing man-child and say my parents cut me off and I can’t afford Comcast anymore.

I always look for the surprising angles in a development of this kind: the knock-on effects. Given 200GW of capacity, in Saudi Arabia, on some days there will be endless electricity.

What I'm thinking is that this will lead to innovation in industrial processes that can productively use a lot of energy in a 4-8 hour window...something you can turn on in a big way and then turn off. Such a process effectively "stores" the energy in the product. The larger the electricity share of such a process, the closer the marginal cost gets to zero during sunny times.

Most modern industrial processes aren't like that: either they're not very power constrained (e.g., a toy factory) or they need the power to be on for a while (e.g., a steel mill). But then again it makes sense these are the types of processing we've perfected, since that's the kind of power we've hard so far in large quantities (coal, nuclear).

Does anyone know of products that can be made quickly by the application of raw energy? I don't know how desalination works, but (just as an example) if there were such a desalination process then Saudi Arabia could desalinate tremendous amounts of water and store it...or even export it to neighbors.

I am not so excited about the idea of a fundamentalist monarchy---one that is helping starve the poor Yemenis---getting even more power, but I am excited about the downstream innovations the fact of nearly-free electricity could potentially create. If SA invents such processes, then they could be copied in other places.

Yes but my point is: they could've just made money for the next fifteen years or so. Then they could've maybe sold the brand, mapping division and data to the company that kills them, or done some sort of merger. Companies don't have to live forever to make sense. Every apartment building, for instance, will eventually need to be replaced.

The self driving car program seems like reaching for immortality. Just because your company identifies something that will eventually kill it on a long enough timeline doesn’t mean you need to develop that product yourself. Wonder how much better off they could have been if they had focused on tweaking aspects of their current service.

This is very true. When I was in grad school I made a group to teach data vis to grad students. The distribution of learning, in retrospect, was super lopsided: almost everyone who did our activities wasted their time; a handful became obsessed and are now professionals who use those skills and many others constantly.

Since interest rates are positive now, The number of jobs is currently controlled by the fed. The stronger labor demand is, the more they raise interest rates, thereby reducing job creation. So all of these make work programs, like the huge military spending boost, are currently just leading to higher interest rates and prices and shifting workers from one industry to another. Specifically it will shift workers out of export production and construction, because higher interest rates hurt our exports and construction. No job creation, more expensive products and homes. Thanks. Fortunately these tariffs are pretty insignificant: 50 billion worth of stuff is not that much for us, although it might be for certain cities in China and the precedent is bad.

This is not what I expected at all. The original news reports based on the police comments made it seem like she was on the side of the road and then randomly darted out. That's not something I would really expect AV's to be able to cope with yet, if ever.

Instead what we see is a scenario that happens all the time: a pedestrian who is sleepy or perhaps with mental problems or on some substances...in any case not hyper alert...crossing a road without looking. It's not that busy at that time of night and I'm sure I have done the same thing. It's very discouraging and angering that these cars are being driven without the capability. I feel very disappointed. To me it's not super important whether an ordinary driver would've stopped: ordinary people drive drunk, exhausted and/or distracted all the time; but I hoped that AV's were already better than that.

so cool they are finally gonna let the parents wander around freely.

This is why I switched from cleaning spray to chewing tobacco.

It does not matter what particular country produces one one's steel. What matters is the price of steel. US mills will be able to charge higher prices, so the cost of NYC construction will rise; although in the short run builders have probably locked into certain contracts. That is the whole point of the tariffs: to raise steel prices.

This is not quite Pareto distributed but it's getting there:

"half of all Airbnb rentals that are conducted by only 10 percent of hosts, who earned a full 48 percent of all the revenue earned in the city last year. That’s some 5,000-people earning a combined $318 million. In contrast, the bottom 80 percent of New York’s hosts—the city’s 40,400 true home sharers—earned just 32 percent of all revenue, or $209 million, in 2017."

While I like the exploration of facts like this, I don't really believe AirBnB is important to rents in New York City. In spatial statistics it is difficult to really prove anything rigorously, and this study doesn't even try to go the distance.

"Applying Barron et al.’s national average ratio of exogeneity to New York City, this implies that, city-wide, Airbnb drove up rents by 0.8% in 2015, 0.4% in 2016, and 0.2% in 2017 (all for years ending in August). This is a cumulative 1.4% increase in rents over these three years attributable to Airbnb’s presence in the city."

So while it might seem like the study looked statistically at AirBnB's impact on NYC, in fact it did not. They just used a sort of rough rule-of-thumb from a national average. But the rule-of-thumb is useless in particular cities (suppose a city had one unit on AirBnB; would adding one more increase rents?), and isn't even from published work. The authors of this study are not economists or econometricians, and it isn't peer-reviewed and never will be.

Moreoever, you can use common sense:

"Airbnb has removed between"

(Dr. Evil voice)

"7,000 and 13,500 units."

New York City has 3.4 million housing units. While year to year housing production varies a lot, on average it produces that number of new units every few months: https://imgur.com/a/gnzUf The report also notes that these units are concentrated in the most expensive neighborhoods anyway. And NYC is only part of a much larger housing regional housing market. It simply strains credulity to think even the upper bound of 13K units has any impact---especially when you consider some of the people occupying such units would have letted or subletted in NYC anyway (the report notes many are long term rentals).

Of course, the real action is in zoning. NYC has seen some upzonings, but mainly massive, consistent downzonings over the past few decades. It is funny people worry about about supply when a private firm removes a tiny bit of it, but nonchalant when the government removes lots of it. Construction costs have also soared, and are now sure to get worse thanks to Trump's tariffs on steel, aluminum and canadian timber.

This whole piece is a good example of how fine writing can hoodwink you. Garbage analysis from start to finish.

"Unlike in Europe, where cities and towns existed long before industrialization, much of America’s heartland has no remaining preindustrial history, given the destruction of Native American societies. The gutting of that industrial backbone — especially as globalization intensified in a country where market forces are least restrained — has been not just an economic fact but a cultural, even spiritual devastation."

From time immemorial, people---especially cultural critics and journalists---have blamed social phenomena on culture, the era's degradation, political changes or something else they are interested in anyway. It flatters you for your command of politics and history: turns out all that stuff you read about because it interests you is actually solidly causative in explaining something important. What a nice coincidence.

I am pretty sure that when the history of this era is written by real social scientists, what they will decide is much more technical and boring...having little to do with the topics that capture our attention for other reasons (manufacturing jobs and trump's election in this case). For example: a bunch of new opioids had been invented, which insurance paid for and doctors prescribed. Large amounts of heroin were available at low prices from increasingly competent and seamless sellers. In what world would these phenomena take place and the amount of addiction not increase?

I used to live in England. There is not as much of a crisis there although there is plenty of economic hardship, loss of manufacturing jobs, etc. The difference is that the NHS won't pay for everybody to load up on a thousand pills, and England is an island where you can't easily smuggle in tons of heroin.

Sometimes the article really veers into the absurd. Consider this passage:

"A huge boom was kick-started by the Civil War, when many states cultivated poppies in order to treat not only the excruciating pain of horrific injuries but endemic dysentery. Booth notes that 10 million opium pills and 2 million ounces of opiates in powder or tinctures were distributed by Union forces. Subsequently, vast numbers of veterans became addicted — the condition became known as “Soldier’s Disease” — and their high became more intense with the developments of morphine and the hypodermic needle. They were joined by millions of wives, sisters, and mothers who, consumed by postwar grief, sought refuge in the obliviating joy that opiates offered."

Right here Sullivan offers a total explanation for why everyone got addicted to painkillers: the government was literally manufacturing them and giving them in massive quantities. But that's not literary enough, so he turns around and fabricates something out of whole cloth:

"the epidemic of the late 1860s and 1870s was probably more widespread, if far less intense, than today’s — a response to the way in which the war tore up settled ways of life, as industrialization transformed the landscape, and as huge social change generated acute emotional distress. This aspect of the epidemic — as a response to mass social and cultural dislocation — was also clear among the working classes in the earlier part of the 19th century in Britain. As small armies of human beings were lured from their accustomed rural environments, with traditions and seasons and community, and thrown into vast new industrialized cities, the psychic stress gave opium an allure not even alcohol could match."

Ah so it was the "psychic stress" of industrialization. So losing manufacturing jobs causes addiction, but also so does gaining them. Interesting that manufacturing jobs are such a powerful explanation in an era when our president is constantly banging on about manufacturing jobs. Suppose we hadn't industrialized at that time; I'm pretty sure people still would have gotten addicted given how many free opiates they were getting. In that case Sullivan probably would have said, "The rural isolation of Americans...the soul-crushing vicissitudes of farming born alone...caused Amercians to take refuge in opiates."

Also worth remembering Andrew Sullivan is serially wrong.

The difference is in the MO for publishing in a given field...not the conscious venue selection so much. From what I know, computer scientists have a high publication rate, rank conference publications as pretty important and sometimes write papers that are about something they built.

Compare this to economists. For economists, it's all about hitting a home run a couple times...publishing a few career-making papers in the best possible journals. Economists will publish only a couple times per year, and their papers are sometimes very long. The review process is crazy: the time between first draft of the paper and final publication can be like five years.

I think what determines this is the difficulty of ascertaining the quality of a piece of work. In CS or math or math physics, I think anyone in a certain subfield is capable of ascertaining how good a paper is. It doesn't need to be extremely long or flashy, and chances are it will make an incremental improvement. By contrast, let's say an economist publishes a paper called "do red light cameras encourage speeding?" The intended audience (economists) doesn't really know anything about red light cameras or driver behavior. They don't have the time to do all the data analyses themselves, and until recently would never even have access to the data. So they look for a bunch of other markers of quality: writing style, citations, length, venue, etc.

Scholars can often make their work free via pre-print. The issue is that the published version isn't free, and the pre-print is hard to find and often not as nice looking.

I made a github repo of my last paper published in an Elsevier journal. This was cool because I could put the latex and simulation code for others to copy. https://github.com/lewis500/distribution-of-trip-lengths

The final version differs a little bit because they had a copy-editor go through and make some good improvements, even after the referees made their comments. Interestingly, the journal was fine with my linking to the repo within the text of the paper, as a way to share the code.

There is nothing like real open access though. This journal has wonderful typesetting and such and is totally free: https://www.jtlu.org/index.php/jtlu/article/view/835/860

It's funny what tricks they use to justify their existence. Always emailing me about bizarre author features no one will use. Makes me sad to think of the engineers sitting in a big room year-on-year polishing features no one will use.