Grade inflation at Ivy League universities

From The Economist

Textbooks and price gouging

From The Economist

In the US, even journalists practice cost benefit analysis

From today's NYT, on the travel ban in NYC declared before the snow storm:

"As of 2014, there were 3.9 million people working in New York City, earning an average daily wage of $409. A majority of those workers commute via the city’s public transit system, even when the roads are in good condition. If the subway closing led just 10 percent of people who work in New York City to take the day off today, the cost in lost labor was around $160 million — lost wages for those who are not fortunate enough to get a paid snow day, and lost productivity to the employers of those who did get paid without working.

Of course, you have to weigh that cost against the cost of being caught unaware in a snowstorm that’s really big. 

(...)

What’s harder to defend is the choice to close the subway, which is more resilient to snow than the roads are, in large part because so much of it is underground. 

(...)

In 2010, an A train got stuck in the snow for 10 hours, stranding several hundred people. The M.T.A.ended up paying a $2,500 settlement to each stranded rider. It was unfortunate (particularly for the riders), but if a stuck train is the worst thing that might happen when running the subway in the snow, that makes a very strong case for running the subway in the snow. After all, those settlements totaled a bit over a million dollars, a small price to pay for keeping the city running and working for a day."

US journalists also understand political economy pretty well:

"Politicians face strong political incentives to overreact to crises. Mr. Cuomo made a similar overreaction to the Ebola crisis several months ago by ordering quarantines of returning medical workers that were widely considered unnecessary and punitive by public health experts.

He wasn’t wrong on the politics; an unnecessary abundance of caution on Ebola polled well. But that doesn’t make it smart policy."

The cargo industry

A very interesting article on the industrial organization (* see bottom of post) characteristics of the cargo industry, here

Excerpts:

Until the late 1990s, the largest container ships could carry about 5,000 steel shipping containers, each about 20 feet long. Today, such ships are little more than chum. The size of container ships has exploded, reflecting their role as the packhorses of globalization. Each year, the maritime shipping industry transports nearly $13 trillion of goods, roughly 70 percent of total freight, according to the World Trade Organization.

The Triple-E’s can carry more than 18,000 containers, piled 20 high, with 10 above deck and 10 below. But they can sail only between Europe and Asia, as their nearly 194-foot wide hull is too large to fit into American ports or to slip through the Panama Canal.

(...)

The Mary — stacked so high with cargo that little is visible beyond two smokestacks and a glassed-in command center — is an apt symbol for an increasingly global marketplace. But it also represents the container shipping industry’s overreaching ambitions. Few carriers besides Maersk are profitable, too many new ships are being built, and demand for space on container ships is slowing as economies in Europe and Asia face headwinds.

(...)

“There’s too much capacity in the market and that drives down prices,” he continued. “From an industry perspective, it doesn’t make any sense. But from an individual company perspective, it makes a lot of sense. It’s a very tricky thing.”

(...)

The industry wants ships that carry more containers, more slowly. Fuel prices are a major factor, so ships now commonly “slow steam” to save fuel, cruising at 16 or 18 knots instead of 22. A typical trip from Poland to China takes 34 days.

During a recent voyage to the Suez, the Mary’s crew sailed on a parallel course with a 10-year-old Maersk container ship that held half as much cargo, but the Mary used only 6 percent more fuel. 

(...)

But fuel is only one part of the equation. “The supply of ships has far outstripped the growth rate” of container traffic, said Richard Meade, the editor of Lloyd’s List, a leading nautical newspaper, adding that the top shipping lines “have entered into an arms race in terms of size, led by Maersk” and its Swiss rival, the privately held Mediterranean Shipping Company. Newer ships, he said, “are more efficient, more economically viable and more environmentally friendly, but they are only going to deliver those results if they are full.”

(...)

The shipping industry has also lacked significant consolidation, with shipping giants often seen as national assets. The top five container lines are either family- or state-controlled. Revenue at Maersk, publicly traded but family controlled, equals more than 14 percent of Denmark’s gross domestic product.

Shipyards, conditioned to build ever-larger vessels, are cutting prices to keep their order books filled. The Triple-E’s were built for $190 million a ship, which in 2011 was seen as a relative bargain. By comparison, in 2007 China Shipping Container Lines, another major shipping line, paid $1.36 billion for eight ships, or about $170 million a ship — but those had a capacity of about 13,300 containers, nearly 5,000 fewer per ship.

“In this down cycle, the new-built prices are low and money is cheap, so you would much rather go and buy the vessels than go and acquire a company” that has older ships, said Martin Dixon,  director of research products at Drewry. “Many shipping lines are struggling to make money, so cost leadership is key to survival. Hence, you’re seeing a lot of investment in bigger ships.”

The bigger ships, though, have been sustained by a growth rate in containerization traffic that has been two and a half to three times global economic growth for decades — and that seems to be coming to an end.

“There are two types of companies that will survive this,” Mr. Sanders of Boston Consulting said. “Either you have the very large companies like Maersk” that “take advantage of scale and make money, or particular shipping lines that operate a niche where they dominate, like a feeder line out in Southeast Asia.

“The other guys,” he added, “are caught in the middle and will have a hard time to make a decent return.”

(...)

While the ships grow, the crews don’t, in another economy of scale. The Mary has a multinational crew of 20 to 30 — and with so few, they multitask.

(...)

The cost of the crew and a range of other expenses related to running the vessel account for more than a quarter of the Maersk Line division’s cost base. Fuel represents more than a fifth. Cutting such costs helps Maersk steady its results.

(...)

Earlier this year, the Nicaraguan government announced plans to build a new canal system, bankrolled by a Chinese billionaire, that would compete with Panama and be wide enough to accommodate the Triple-E. Such plans have been floated before and failed to materialize, but it is the kind of audacious project that only the Chinese might try.

No country more than China has spurred the containerization boom, a byproduct of moving the world’s factories thousands of miles from their biggest markets.

(...)

China is also exerting its influence on the industry. The country’s regulators recently blocked an attempt by Maersk and two European rivals to form a partnership, saying it would hinder competition. Maersk recently proposed a less ambitious alliance with a single rival, the Mediterranean Shipping Company. The proposed deals represent a concession that cargo volume is not rebounding as quickly, so shippers need to share costs and cargo space. Boats are typically full heading west to Europe and partly empty heading east to China. Cargo from Europe to Asia has grown about 30 percent in the last five years, in part because of rising demand from a growing Chinese middle class. But it has not nearly filled all the containers.

At the same time, some manufacturing is moving closer to local markets, a trend that contributes to slowing growth in container traffic. China, too, is trying to foster its own shipping lines. This year, China Shipping Container Lines ordered five ships that will each hold 19,000 containers, about 1,000 more than the Triple-E. They begin to arrive later this year."

(*) If you wonder what industrial organization is, let me quote Uncle Bernard (in French):

"Les théoriciens de l'économie industrielle sont une secte, dont l'obscurantisme et le fanatisme donnent froid dans le dos. Il n'est pas difficile de repérer le taliban sous l'expert, et le fou de Dieu sous le fou de l'incitation."

So, you will I hope excuse me if I am NOT Bernard...

À Montréal, sous -30 degrés, nous sommes tous Charlie

L'impact des aides au logement sur le secteur locatif privé en France

L'INSEE a publié un document en novembre dernier sur "L'impact des aides au logement sur le secteur locatif privé en France". Sa lecture est édifiante. On y apprend notamment que:

- Ces aides représentent un montant de dépenses publiques très important, et en forte croissance depuis les années 70:

"En 2012, les pouvoirs publics ont consacré 15,8 milliards d’euros d’aides personnelles pour les locataires dont 8,1 milliards pour le secteur locatif privé. Les aides au logement regroupent l’aide personnalisée au logement (APL), l’allocation de logement familiale (ALF) et l’allocation de logement sociale (ALS - sources et méthodes). L’augmentation du budget de l’État consacré à ces aides est continue depuis la fin des années 1970 et a été particulièrement forte pendant les années
1990 (figure 1). Cette augmentation résulte d’un fort accroissement du nombre de bénéficiaires au début des années 1990 et d’une hausse du montant des aides par bénéficiaire. En 2012, 5,7 millions de ménages percevaient une aide au logement."


- Et pourtant, leur efficacité n'a pratiquement jamais été étudiée!!! Les deux seules études citées pour la France datent de 2002 (Laferrère et le Blanc) et 2005 (Fack). Vu le contexte actuel des finances publiques, il est pour le moins surprenant qu'un volet aussi important des dépenses publiques ne fasse pas l'objet d'une évaluation approfondie et régulière...


- Le document de l'INSEE résume clairement objectifs et effets non recherchés de ces aides:

"Ces allocations ont plusieurs objectifs : limiter le taux d’effort des ménages locataires bénéficiaires ou leur permettre d’accéder à des logements de meilleure qualité, à taux d’effort donné. Néanmoins, ces aides peuvent avoir un effet inflationniste : comme elles permettent à certains ménages d’accéder à des logements de meilleure qualité, la demande s’accentue et peut conduire à une hausse des loyers si le nombre et la qualité des logements ne s’ajustent pas suffisamment. En d’autres termes, une partie du bénéfice de l’allocation des ménages serait alors transférée aux bailleurs. Cet effet inflationniste serait d’autant plus fort que les bailleurs ont la capacité de connaître les locataires éligibles et perçoivent parfois directement ces aides."

- La méthodologie utilisée par l'INSEE pour cette évaluation est la suivante:

"Pour évaluer l’efficacité des aides, il est donc important de quantifier leur impact sur le montant des loyers, sur la qualité et sur la quantité des logements locatifs offerts. La méthode d’évaluation utilisée repose ici sur les discontinuités géographiques dans le versement de l’aide, c’est-à-dire sur l’existence d’un zonage qui affecte le montant des aides versées. Ce zonage a été établi en 1978 et a été très peu modifié depuis.
(...)
 La zone I inclut l’agglomération de Paris et les villes nouvelles franciliennes. La zone II regroupe toutes les agglomérations qui comprenaient plus de 100 000 habitants, les franges de l’Île-deFrance ainsi que quelques agglomérations plus petites et présentant un marché immobilier tendu (par exemple les zones frontalières et côtières) ou durement frappées par la crise des années 1970. La zone III correspond au reste du territoire, c’est-à- dire la grande majorité des agglomérations de moins de 100 000 habitants en 1978. Le montant des aides est plus élevé en zone II qu’en zone III.

Aussi doit-on chercher à mesurer toutes choses égales par ailleurs les écarts de loyers entre la zone II, où les aides sont plus élevées, et la zone III. À cet effet, on comparera donc les agglomérations dont la population est juste inférieure à 100 000 habitants avec celles situées juste au-dessus de ce seuil, en contrôlant des caractéristiques observées des logements et des agglomérations"

- Les résultats obtenus sont

(1) Les propriétaires captent (une partie de) l'aide destinée aux locataires :

"Les résultats obtenus par cette méthode indiquent que les loyers sont significativement plus élevés dans les agglomérations de plus de 100 000 habitants, suggérant un lien causal entre les aides au logement et le niveau des loyers privés. Cette étude confirme et actualise ainsi des résultats établis en France et dans d’autres pays".

Malheureusement, " La méthode utilisée ne permet pas de calculer la part de l’aide absorbée par la
hausse de loyer."

(2) Ces aides n'ont pas amélioré la qualité des logements:

"Les estimations indiquent que les aides plus importantes dans la zone II n’auraient eu aucun effet sur la qualité des logements locatifs privés, telle qu’elle peut être mesurée dans les enquêtes"

(3) Ces aides n'ont pas augmenté le nombre de logements offerts à la location:

"De même, l’augmentation des aides liée au zonage semble n’avoir aucun impact sur le
nombre de logements locatifs offerts, tel que mesuré par la part du secteur locatif
privé dans la commune. L’effet sur le nombre de logements, et non pas leur part,
est également non significatif."


- Ces résultats sont en fait peu surprenants, et comme indiqués au point (1) ci-dessus ils confirment les études déjà effectuées, en France et surtout à l'étranger.

Ma conclusion: des aides dispendieuses, qui n'atteignent aucun de leurs objectifs fixés, mais qui non seulement perdurent mais coûtent de plus en plus cher aux pouvoirs publics. Mais apparemment, ce bilan calamiteux ne suffit pas à les mettre en cause, même dans le contexte budgétaire actuel. Un beau thème de recherches en économie politique...

3 graphs relating polarization and inequality in the US

From Eric Voeten:

John’s post on polarization and broken politics is great. One thing to add, though, is to link polarization to the issue of the day: inequality in wealth and income. Nolan McCarthy, Keith Poole, and Howard Rosenthal (my dissertation mentor) document in Polarized America how inequality feeds into polarization and how polarization, in turn, sets the stage for policies that further increase inequality.

Their evidence comes both from surveys (i.e. how individuals vote based on their incomes) as well as from looking at long swats of American history. The graph below tracks the Gini index, a widely used measure of economic inequality, and the polarization measure John used. The correlation is staggeringly high.  In the post World War II period, they find an almost identically high correlation between the share of income earned by the top one percent and polarization.

Patterns in polarization also move in tandem with the percentage of the American population that is foreign born. Immigrants tend to be disproportionally poor and less likely to (be able to) vote.  If the lower portion of the income distribution contains large numbers of non-voters, then this reduces pressures for redistributive policies that might decrease inequality.

Finally, polarization seems to move with trends in education and wages in the financial services relative to other industries (see below).

I took the graphs from Keith Poole’s page, which has more info on data and background. The causal stories are not always clear and clean. Surely the various social and economic trends reinforce each other. Polarization may well be both a cause and consequence of all these trends. A simple blog post cannot do justice to these complexities. Nonetheless, I wished that the current debates about broken politics focused more on these bigger socio-economic questions and less on immediate causes.

Funding scientific research

The New York Times has a nice article on how a Dartmouth physics professor has built a supercomputer using PlayStation consoles, to do research 

"focused on finding and studying gravitational waves, vibrations that ripple through space-time. The waves, first predicted by Einstein’s theory of general relativity, form after a particularly violent astrophysical event, like two black holes smashing together. Because black holes cannot be observed through telescopes, Dr. Khanna uses supercomputers to create simulations of these collisions."

What I find most interesting is how he financed his research. 

"Because the National Science Foundation, which funds much of Dr. Khanna’s research, might not have viewed the bulk buying of video game consoles as a responsible use of grant money, he reached out to Sony Computer Entertainment America, the company behind the PlayStation 3. Sony donated four consoles to the experiment; Dr. Khanna’s university paid for eight more, and Dr. Khanna bought another four. "

(He then publishes his research results obtained with this 16-Playstation supercomputer)

"Dr. Khanna’s observations caught the attention of the Air Force Research Laboratory in Rome, N.Y., whose scientists were investigating PlayStation 3 processors. In 2010, the lab built its ownPlayStation 3 supercomputer using 1,716 consoles to conduct radar image processing for urban surveillance. “Our PS3 supercomputer is capable of processing the complex computations required to create a detailed image of an entire city from radar data,” said Mark Barnell, the director of high performance computing at the Air Force Research Laboratory. The lab later entered into a cooperative research-and-development agreement with Dr. Khanna’s team, donating 176 PlayStation 3 consoles."

So, for a total of 176+16=192 consoles, the financing is 

0% by publicly funded research funds,

4.2% by the University

2.1% by the private sector

2.1% by himself

91.7% by the military

This tally looks a bit scary to me ...