Saturday assorted links

by on November 11, 2017 at 1:04 pm in Uncategorized | Permalink

1 rayward November 11, 2017 at 1:24 pm

3. “States like Ohio, West Virginia, Michigan and Illinois have been among the hardest hit, with retail employment declining over the past decade, and now those woes are likely to spread. Many states, such as Nevada, Florida and Arkansas, have overly relied on retail for job growth, so they could feel more pain as the fallout deepens.” I suppose this relates to Caplan’s new book: the education system is preparing graduates for little more than a career in retail, while retail jobs are disappearing. But there’s a lot more to the decline in retail than employment, as retail has been the path to both the middle class and business ownership. I recall the time when I was growing up in a small town that hardware stores, clothing stores, grocery stores, and so on were locally-owned family businesses, with each owned by and employing several generations of family members who enjoyed at least a middle class lifestyle. In the future will all their heirs become bankers, computer programmers, and quants, or will they be added to the legions of formerly middle class suffering a decline in prosperity?


2 Ted Craig November 11, 2017 at 1:27 pm

3. I’m glad they make the point that working in a distribution center is not the same as working in a shop.


3 Al November 11, 2017 at 2:05 pm

No doubt. It is tremendously more efficient.


4 a clockwork apriori November 11, 2017 at 8:37 pm



5 a clockwork apriori November 11, 2017 at 8:44 pm

oops, andy the answer id is def 7234, which if considered as Fahrenheit, then the kelvin is 4274.26111 invoking Casablanca e^2 hemingway third man theory


6 reply to the Burgess fan November 11, 2017 at 9:46 pm

Hemingway suffered from CTEs. Welles, in one of his last interviews, stated that almost every writer of his day was influenced by Hemingway ( he was not impressed by those who were so influenced, and he wished it wasn’t true, but that is neither here nor there). Graham Greene, who scored some good paydays from that movie with the zithers, was funny in person but nobody reads him anymore because he did not understand the world. Reality is of course very important (Remember the first time you travelled down a night-time street – walking in the cold, sitting in the back of a car, or, if you are lucky, you remember that day, hand in hand with your first date, coming home from the homecoming dance) and, for example, someone who cares about reality will remember exactly how the street lights looked that night, that dark night when one was so happy. As famous and rich as they were, Ernest and Graham never remembered. That night-time street. I remember.

7 reply to the Burgess fan November 11, 2017 at 9:47 pm

And – here is the best part – the street lights remember, too.

8 reply to the Burgess fan November 11, 2017 at 9:52 pm

Liber Psalmorum, Enumeratio Graeca, 89:5

9 reply to the Burgess fan November 11, 2017 at 9:59 pm

Tehillim 89:5 (Shomayim! as if every streetlight of every memory you ever had were, in blessed fact, what it was created to be: think about that)

10 anonymous reply to the Burgess fan November 11, 2017 at 10:07 pm

Tehillim 89:5

11 a clock work's a prior griot function November 11, 2017 at 11:23 pm

Yes, often I see a row of greenlights and harken back to a time along fulton street when it was the green lights and some scattered trash, delivery bags, of food half-eaten and thrown like spring furniture onto a street, the poor bags stuck to the food, unable to free themselves, and the food itself rotted like an artery so the rats from east river park won’t even make their way across the Williamsburg bridge and down a mile or so to enjoy what a rat, one would think would find worth the travel, and yet still one might expect a delivery boy on fulton street at 3 am on a Saturday night.

12 anonymous reply to the Burgess fan November 11, 2017 at 11:56 pm

thanks. only twice in my life have I had a companion (for most of us, there is only once) who was for me not another person but “her heart and my heart were one heart” and we watched Ratatouille, not that good a movie, by the way, but we both understood – those streets, with those lights, even unpleasantly greenish, with the happy rats just trying to live a happy life as a rat scurrying somewhere almost out of sight lives its life, like any other sinner, hoping for a better day – well, I can picture your comment, as if it were painted by Utrillo on a good day. (For the record nobody – not even the rats themselves – wants rats to be more than liminal figures in any urban landscape. Do you remember the autumn streetlights of the early 1980s – Paris near the river Hudson, Albany near the river Seine, NYC near the cold Pacific? – of course you do.)

13 Bob November 11, 2017 at 1:50 pm

6 Covers interesting things about tokens,(it’s rare to see an explanation discussing regular accounting ledgers, which have most of the properties of cryptocurrencies already) but there’s also quite a bit of cryptocurrency related BS that isn’t necessarily correct, or even handed.

What makes cryptocurrency tokens interesting across digital tokens is that they are transferable, can represent a lot of value, and have absolutely no way to deal with theft or copy. In the real world, we limit the value of individual, anonymous tokens (like, say, stamps), because of risk of theft. It’s also why storing a lot of value under a mattress is not ideal, and we instead put money in banks. All other high value digital tokens out there, like the private certificate that might identify google or encrypt credit card vaults, are protected by layers and layers of security, and they also have mechanisms to invalidate them after the fact in case of theft: If I steal an issuing RSA certificate from a registrar, or fnd the cert hat would let me impersonate Google in front of Chrome, the world would route around me, invalidating my tokens quite quickly after the copy is discovered. In a cryptocurrency it’s still technically possible, but it creates a terrifying event that negatively impacts everyone, and will only be done if a small oligarchy thinks it’s good for them (i.e, they will lose a lot of money themselves). It also highlights that said oligarchy can make a mockery out of said trustless system, because it shows they can break the rules whenever they feel like it.

Many people that are bullish regarding cryptocurrencies don’t take look at this interesting property, and therefore ignore major practical downsides. There’s also the throughput problem, which is why you find very few distributed systems experts betting on cryptocurrencies, but that’s for another post.


14 Ricardo November 12, 2017 at 3:34 am

Right, Bitcoin introduces non-trivial risks and costs to its user community by being pseudonymous and relying on distributed databases. If you take those two features away, you have the existing global banking system which works well enough for the vast majority of law-abiding people. Citibank, HSBC, Chase, etc. don’t rely on decentralized blockchains to keep track of customer accounts and deposit insurance, regulation and the court system almost certainly provide depositors with more security than Bitcoin does.


15 albatross November 13, 2017 at 12:06 pm

Blockchains let you keep records in a public, unalterable way, where the rules of how things are added to the chain are public and the power to do so isn’t concentrated in just one organization’s hands. Cryptocurrencies are something you can do with a blockchain, as with Bitcoin or ethereum. The existing banking system can probably benefit in terms of security from blockchain technology, even though it wouldn’t make any sense for them to try to move all their transactions to Bitcoin or something.


16 Matthew Young November 11, 2017 at 2:02 pm

The economics of tokens (long, interesting, and dizzying).

Bitching for six months that Fintech did not separate the semantics and let everyone be fooled by ICO. Finally this author steps up to the plate. Not all tokens are a unit of account, like a coin should be. Instead of ICO we need ITO, initial token offering, then let the contract tell you what kind of token.

Make the contract operational by your trading bot. There is protocol theory that let’s bots understand token protocols.


17 Paraguayan November 11, 2017 at 2:04 pm

In America there is no limit to how far a man may fail upward.


18 derek November 11, 2017 at 2:42 pm

1. Didn’t the Democrats structure their primary system to prevent that from happening, with the super delegates not tied to any primary result? Which resulted in Hillary, probably the only Democrat who could lose to Trump. Trump approached the process as it was; all he needed to do was beat the field, which he did. Same in the election, he needed electoral college votes, so he got them.

3. The Federal Reserve policy of almost free money both encouraged this and delayed it’s outcome. I was talking to someone who did grocery store construction, and he said that store chains would buy land early and build during slow times when prices were reasonable. That changed in the 00’s, where the per sq ft construction costs were very high. Money was available and cheap, and instead of competing on the price of bread and milk, they were competing on how quickly they could build 150000 sq ft stores. It seems to have finally caught up with them. I’m watching one of my customers go down the spiral to it’s inevitable end; they are squeezing costs and optimizing to the point where the deli is a pig sty, no one is facing the shelves, and items that people buy are no longer on the shelf.


19 Emmett November 11, 2017 at 4:38 pm

The Condorcet Paradox on shifting majorities is irrelevant to the Democrat and Republican primary system … where well under 20% of the ostensible electorate ever votes for anybody — a dart board would be more fair


20 ChrisA November 12, 2017 at 1:13 am

@Derek – you said “The Federal Reserve policy of almost free money..” – you do know that what matters is real not nominal interest rates? Look up the term “money illusion”. Real interest rates are not particularly low. If you want higher nominal interest rates, then you will need to generate more inflation. The Fed would be able to do that by increasing money supply.


21 Art Deco November 11, 2017 at 4:22 pm

1. It is difficult to believe this is true globally. The sum of ballots cast for Messrs. Cruz, Rubio, and Kasich exceeded ballots for Trump by only about 10%. You eliminate Kasich and redistribute his balllots to the 2d choices among the remaining 3, eliminate Rubio and redistribute his ballots to 2d and 3d choices among the remaining two and it’s difficult to see how you arrive at a situation where Trump is outpolled by the remaining candidate unless you assume that he is no one’s second choice or third choice. (And for those content with the Capitol Hill nexus, Cruz is as unpalatable as Trump, albeit for different reasons).


22 shrikanthk November 11, 2017 at 5:46 pm


I don’t see Trump as the Condorcet loser in the primaries either. Maybe Rubio had the best chance in a head-to-head contest


23 EricCh November 11, 2017 at 5:15 pm

#4: India is number 1 for sure in term of overconfident. A humble Indian is as rare as a boastful Japanese.


24 anonymous reply to EricCH November 11, 2017 at 6:30 pm

I’ve known a couple, from several different castes. But then I am old: I can still remember when the official state song about Virginia positively referenced how happy the lower castes were in the Old Dominion. I imagine that those born after, say, 1980 would not have noticed this, unless they an exceptional gift for cross-cultrual communication.


25 November 11, 2017 at 7:59 pm

#4 TL;DR but the results might be clouded by diff pops affinity to maths, gambling and risk taking.

I have my own ‘index of underconfidence’ at the pop levels derived from the OECD PISA data. The confidence index is defined as the ratio of OECD PISA score over the pct of pop who think they are the best. As such positive value means underconfidence. The values are normalized with the sample mean and stddev to give z-scores. Underconfidence if the index is > 0.5 SD and overconfidence if that is < -0.5 SD, otherwise they are ‘normal’.

Rank Index Country


1 3.09 Netherlands

2 2.73 Japan

3 1.67 Switzerland

4 1.5 Finland

5 1.4 Belgium

6 1.28 Germany

10 1.06 Macao_China

11 0.99 France

15 0.71 Estonia


26 November 11, 2017 at 8:00 pm

—– (normal)

16 0.46 Italy

18 0.27 Russia

23 -0.04 ChineseTaipei

24 -0.09 Norway

25 -0.1 Sweden

30 -0.23 Denmark

31 -0.27 HongKong_China

32 -0.31 Greece

33 -0.34 NewZealand

34 -0.34 Canada

35 -0.37 Ireland

36 -0.41 Singapore

37 -0.47 Australia


27 November 11, 2017 at 8:01 pm

—– (overconfidence)

39 -0.51 BSJG_China

40 -0.53 UK

42 -0.56 Korea

44 -0.73 Brazil

46 -0.86 US

47 -0.89 Israel

49 -1.03 Mexico

50 -1.13 Turkey

51 -1.14 CostaRica

52 -1.15 UAE

53 -1.2 Qatar

54 -1.24 Peru

55 -1.28 Colombia

56 -1.39 Tunisia

57 -1.5 DominicanRepublic


28 November 11, 2017 at 10:57 pm

It is strange that the underconfidence and overconfidence groups share the same statistically significant trend where the level of confidence drops with increasing IQ, i.e. those that dont know what they dont know and those that know what they dont know, while for the normal group they just dont care (the trend is not statistically significant) and a high fraction of them have the ‘grit gene’.

In a way it could be argued that the defined normal range is too narrow to show any trends. The data can be split into four quadrants with the sample averages as the origin. Most of the ‘normal’ are in third quadrant and still shows not statistical trend. Thus the general results still stand.

Again it is this finer grain ‘pitch fork’ behaviours and their limited experimental ranges that confused the authors of the cited paper above who concluded that there were no ‘universal trends’. It can be seen from
(on a no registration site, avoid if you are easily triggered)
that their pops from Canada, Japan and HongKong formed a tight range, and Japan is on a different fork than that for Canada and HongKong with opposite trends.


29 stephan November 11, 2017 at 9:47 pm

#6 Digital currencies. This idea that we’re better off ( as a store of value and medium of exchange) with something that is very energy intensive , has high transaction costs, slow transaction throughput and a volatile and unpredictable value makes no sense to me.

I’ll stick with the dollar.


30 Hazel Meade November 11, 2017 at 9:55 pm

3 is contradicted by last Sunday’s 6.


31 ChrisA November 12, 2017 at 1:22 am

On the death of retail – isn’t this just the commodification of clothing and not really an amazon effect? Malls and department stores mostly sell clothing. But supply chains in clothing are so efficient nowadays that prices are very low even for “designer” clothing. As a result profit margins have to be very high to pay for the retail infrastructure in department stores, so cannibalisation by lower overhead specialist stores is just too easy. Probably this is also promoted by increasingly more casual workplaces. Very few work places require suits, ties and dress shirts (or same equivalent for women) anymore.

Overall though I would bet that the retail workforce will continue growing, but more into the restaurants and bars.


32 Albigensian November 13, 2017 at 11:46 am

Some clothing is commodity, some is fashion. And so-called “fast fashion” (low-cost but trendy) continues to sell, at least to some shoppers. The challenge is not just in finding what’s trendy, but in managing a product with a short profitable lifespan that’s sourced through lengthy supply chains.

For non-commodity clothing, physical stores should have an advantage as one can accurately see the colors and feel the fabric (difficult/impossible on a screen), and easily return it if it does not fit.

On the other hand, some retail does resemble “dead men walking.” Does anyone really expect Sears Holdings to survive? Why would it? And the big department-store chains, when department stores are in a long-term decline that predates online retailing by decades?


Leave a Comment

Previous post:

Next post: