Wednesday, March 31, 2010

The cost of Security

When is too much security a bad thing?  One often hears politicians saying "one death is too many", but how does that translate in the real world.

Taking a real world example:  The U.S. government expanded its sky marshal service following the catastrophe of 9/11; today more than 4,000 sky marshals ply the "friendly skies".  Between the end of December 2001 (when the program was first implemented) and December 31, 2009 total arrest by sky marshals was 4, (not per marshal -- in total).  Cost of each arrest was therefore $200 million.

The question is was that an appropriate resource allocation for the government.  But consider this, a number of American shun the airline industry after 9/11 (both fear and additional check-in procedures), the result has been estimated that an additional 3,000 road deaths occurred because of the switch by many Americans away from aircraft (very safe) to their cars (not so safe).

One could say that a percentage of the additional road deaths were caused by the additional inconvenience of flying following the implementation of additional securities measures.  So in fact, the additional security in airports and aircraft achieve the exact opposite of what was intended (e.g. reduce the overall number of deaths).

The law of unintended consequences rules!

Saturday, March 13, 2010

Manufacturing in China -- Not tested on Animals

I just finished reading Poorly made in China by Paul Midler, a short 241 page book that takes a look at manufacturing in China since 2000.  It is very much an anecdotal look at the country, but it also reveals certain aspect of the behavior of the Chinese people.

My favorite bit is halfway through the book, where the American buyer has changed the information at the back of the bottles of shampoo to indicate that the shampoo was not testes on animals.   I can see many people who would find this an admirable situation for a company to commit not to test its products on animals.  The problem is that the product was not tested at all (excluding consumers)!  In a nutshell, the manufacture was continually  changing the formula without consultation with the American buyer (or for that matter telling anyone) without anyone testing the product to see if it was dangerous when exposed to users.

The more serious aspect of desintermediation is here further illustrated; in the financial sector (in the US) those who wrote the mortgages, those who packaged the mortgages into security products and those who financed the mortgage backed securities were three entirely different groups, that had conflicting objectives, the result was the MBS and Alt-A fiasco that has and continues to destroy financial institutions the world over.

In the manufacturing side, it became more convenient (and cheaper) for American soap sellers to get the stuff manufactured in China, where there are no standards or regulations, dramatically reducing the marketing wrath of Greenpeace faced by the American companies' laboratories in Connecticut -- in essence the law of unintended consequences!  Whereas when Greenpeace and other animal rights movement were picketing or firebombing laboratories that conducted quality control test on US manufactures products, they never conceived that the end result would be for more and more of the chemical products we use everyday to be manufactured in place were there are no quality controls checks -- the consumer is at the mercy of the lowest bidder.

The problem left unspoken here is that over time, its not only the cheap stuff that is manufactured in China but also the expensive stuff, because the cost advantage for marginal supplier is even more tantalizing for larger players, in fact large players are active in China using an army of quality control inspectors, yet Mattel found itself caught out -- and that's one that was well publicized.  The risk is that debased goods will flood the market.

Tuesday, March 9, 2010

American Governmental contribution to GDP growth in Q4/2009 was negative

One of my favorit economic blog: Worthwhile Canadian Initiative published a comparison between Canadian and American GDP growth for the Q4/2009 period, comparing Canada's 5.0% GDP growth to America's 5.9%.  Stephen published the table below which compared the source of growth in the two economies:
The third column of the graph has serious implications; underlining the different impact that government spending are having in both jurisdictions.  Whereas the source of GDP growth in Canada was consumption, government spending and investments, in America GDP growth was caused by inventory, consumption and investments.  It appears that  the massive federal government capital injection into the American economy was completely negated by the cuts in expenses by state and local governments.  The overall GDP growth contribution by all levels of government was negative -- the implication of the ending of federal stimulus will therefore become a serious drag on the economy.  A fun fact is that unlike Canadian provinces, American states are constitutionally prohibited from running budget deficits (essentially the American system has installed a system that Europe should have adopted).

The stories one hear from states like Nevada -- where even if the state fired every single employee (we are talking everybody -- teachers, road crew, police, firemen etc) they would still register a budget deficit.    

I'm not entirely clear why so few commentators discuss this aspect of GDP growth.  I was surprised that this was not a talking point on FOX (no not really) -- imagine Fox news stating that government contributed negatively to GDP growth.

Monday, March 8, 2010

More proof of Canada's two solitudes

This morning while reading the papers I discovered that Canada has two different and distinct national anthems.  One in French and one in English.  This is not about the difference in terminology, the two anthems (sung at hockey game every Wednesday night) actually have very different thematic.

The French version is a poem written by a guy called Adolphe-Basile Routhier in 1880.  It remains today largely unchanged from its original text.  The English version is "regularly revised" by the Federal parliament.  What is truly amusing is that as a bi-lingual Canadian I never eve noticed the different texts.  Of course nobody ever talks about this, the only reason its a topic of conversation is that last week in the "Throne Speech" the governor general (Queens representative to the Canadian Government) mentioned that the government was thinking of modifying the English version to remove "Sons" and add something more inclusive and less gender specific.

Nothing really big here, except that the Conservative "electorate" decree this as a crime against humanity.  How dare the government changing the words of the Canadian anthem, it has remain unchanged now for nearly....15 years.  Needless to say that the government back down from this idea as fast as you can say:  Pussywipped!

Anyway, another example of the two solitudes:  We have two national anthems in Canada, they don't talk about the same thing, but they rhyme.

Tuesday, March 2, 2010

The challenges of health care

Yesterday I read a fascinating article written by Atul Gawande, an American surgeon who wrote in the New Yorker magazine on cost control in the U.S. medical system.  His premise is that independently of provider (Government, HMO, other) the issue is about the type of incentive given to the medical establishment. 

His layman analogy was that instead of paying a contractor to build your house you paid the trades; plumbers and electricians on the number of faucets and outlets in your house – the outcome is obvious, too many faucets and electrical outlets.

Gawanda focused on what the most expensive and most cost cautious providers were doing.  His finding (as the above analogy indicates) that the compensation system in Medicare is skewed towards quantity and not quality.  What makes his analysis interesting is that in Texas the cost difference per participants in two regions was dramatically different:  $7k for El Paso and $15 for McAllen.  It is noteworthy that we are talking Medicare here, not HMO or other forms of insurance.  The payer was the same in both cases, with very similar health concerns (all that TexMex goodness…).  As an aside Gawanda indicates that in 1992, both regions had similar costs per participant of $4k.  Something happened in McAllen or El Paso or both!

Read the article it’s here.

The overall issue of interest is not the story itself, but rather the debate is about what healthcare is about.  Places like El Paso and the Mayo Clinic view their role to be greater than just providing specific services, but considering the patient as a whole (we’re not talking touchy feely stuff here, but hard core analysis of what make health care work, and what is meant by good health care, as Mr. FitN often points out “More is not always better, some time less is more”).  The problem is that tests and procedures carry their own additional risk.  Medicine is not like a bigger car.

The problem is not isolated to the U.S. somewhat as a parable: In 2005 one of Canada’s premier teaching hospitals had budgeted 700 stents – this popular and non-invasive procedure (instead of “cracking the chest”) became the preferred method of reducing angina and other blood flow related heart procedures.  Part way through the year, the Hospital’s board became aware that their stents budget was inadequate!  At the end of the year instead of 700 stents the cardiology department had used nearly 1,400 at an additional cost of $420,000. 

The most expensive piece of medical equipment is a doctor’s pen. And, as a rule, hospital executives don’t own the pen caps. Doctors do.

The surgeons had decided to install two stents instead of one.  The board’s reaction was to ask these surgeons on what study they had based this decision.  It turns out there was no study.  It was an unspoken agreement that two was better than one.  The implication is that again the method by which healthcare is delivered is somewhat irrelevant; what is important is to control cost and optimize patient well being (by the way it turns out that installing two stents instead of one has no impact on the patients’ well being or incidence of return to the hospital)..