2009-02-28

Thoughts on doctors not spending much time per patient

"Even as an expert, it takes me over 5 hours to review a typical middle age person’s medical history (plotting test results, comparing to distributions of patients, etc.). To select and optimize the best set of sequential diagnostic tests and treatments takes many hours per patient. No physician spends the time needed. In my review of thousands of cases with chronic conditions, more than 95% provide suboptimal care (excluding minor accidents, simple infections, simple cases).
Conclusion. Today we can spend over $1M per person to improve morbidity and mortality. And tens of hours of professional time. Clearly impossible. Therefore, health care implies rationing and restricting physician’s time, diagnosis and treatment.
Thus, the first priority is to admit that it is impossible to provide the state of the art care to every person. Not even a small proportion of the population. Physicians today allocate less than 10’ per patient, on average. In this time it is impossible to review test results, select optimal sequential testing from a menu of 10K +, order them, review results, and so on.
Once we admit that we can only provide a tiny fraction of what modern medicine provides, then we can optimize the allocation assumming that we can agree on outcomes."
— Eduardo Siguel, MD, PhD
Comment six on The Mounting Price of Health Care’s Status Quo

Here is what I posted in the comments in response:

Thank you Dr. Siguel (Comment 6) for sharing some important insights and providing some validation for something I have been wondering about for years.

For most of my adult life I have been puzzled by how doctors routinely make major decisions with little or no research or analysis. Time after time I saw a doctor make a decision on a complex situation without cracking a book (or a database in this modern age) or reviewing records in detail. I suspected that these decisions were of sub-optimal quality, but hoped (since the health of me and my loved ones was on the line) that somehow these doctors had discovered a way to make complex decisions without doing any work. Dr. Siguel's insider view seems to confirm that that quick medical decisions do tend to be sub-optimal.

However, I have to disagree with Dr. Siguel about the implications of limited time leading to sub-optimal medical decisions. He feels that the solution has to be rationing of doctor's time because providing the time needed for optimal medical decisions would be prohibitively expensive. But what about lowering the cost of doctor's time?

The elephant in the room is that physicians make a lot of money because the supply of physicians is constrained by very high educational and training requirements. Sure we want all our doctors to be genius superstars. But here is the conundrum: what is the use of requiring our physicians to have such high qualifications if the resulting high cost of physician's time means that they mostly shoot from the hip and frequently make sub-optimal decisions? If we are willing to accept sub-optimal decisions, why not at least get them for cheap from less qualified, and therefore less expensive, doctors? If we lower the training requirements for becoming a physician, and open dozens of new medical schools, we could get physician's pay down to say the high 5 figures (as versus the current mid to high 6 figures for specialists), and then it would be affordable to have a physician spend more than 10 minutes per patient.

I personally would jump at the chance to be treated by someone with half, or less, of the qualifications of my current physicians if that person would spend some real time carefully examining my records and doing some research on the options. I don't want a genius, I want a workhorse.

Also, with the advent of computers and data technology, the need for spectacularly trained physicians is declining. In the olden days a physician pretty much had to have a whole library in his head because researching stuff involved, well, going to the library. These days the latest medical research is only a few clicks away, and there is not so much need for a physician to know everything.

2009-02-20

Comment not accepted on Krugman blog

Here is a comment I posted in response to this post on Paul Krugman's blog that apparently didn't make it past the moderator:

http://krugman.blogs.nytimes.com/2009/02/18/the-eschatology-of-lost-decades/

I think its fascinating that so far nobody has really made any predictions for how it will end. So, I will give it a go.

To me the key word for understanding where we are headed is “unsustainable.” I think everyone agrees that the levels of borrowing and spending in the US (both public and private) during the past few years were unsustainable. One way or another the US is going to eventually end up back at sustainable levels of borrowing and spending, and that sustainable level is going to necessarily mean a lower GDP per capita than we had in 2007. Once the economy falls back to a sustainable level of GDP it will start to grow again slowly. So that is my prediction: the economy is never going to “recover” to where it was before, because where it was before was unsustainable.

What is a sustainable level of GDP? I have no idea beyond that its going to be lower than the peak of the bubble.

I see a couple paths back to a sustainable level of consumption:
1. Do it the old-fashioned way and let the contraction run for a few years until consumption bottoms out on its own and all the excess durable goods are used up.

2. Use hyperinflation to wipe out all debts and all financial assets, effectively rebooting the whole economy. Even with hyperinflation we would still have a huge surplus of durable goods (ever noticed how many storage units there are out there?) that we would have to work through, but at least people wouldn’t be investing massive amounts of energy into maintaining an increasingly tangled web of wacky financial arrangements.

3. Have another World War II to blow up all the excess durable goods.

4. Lift all immigration restrictions for people with a positive net worth and a college degree and let all the new immigrants soak up the excess durable goods and boost consumption spending (nothing would stimulate like 10-20 million people with money in the bank setting up house in a new country.) We could kill two birds with one stone by lifting all immigration restrictions for nations facing civil unrest or ethnic conflicts and providing them a subsidy to relocate to the US. That way all the people in Somalia, Gaza, the West Bank, Iraq, Afganistan, the Sudan, etc. who just want to live a peaceful life have a chance to leave conflict behind. When the warring factions see everyone leaving they will realize that if they keep fighting they will soon have nothing left to fight over.

5. Use export sales to replace lower levels of domestic consumption (not looking likely as noted by Prof. Krugman).

I am no expert in economic history, but I cannot think of a single major recession or depression whose recovery didn’t involve one or more of these. So my prediction is that it will end with some combo of the above, just like every other major recession has.

The one thing I am pretty sure of is that bailouts and stimulus packages are not going to magically make unsustainable sustainable.

2009-02-19

A modest proposal to save the economy, regain American dominance, and bring peace to the world

Right now the two things dragging down the US economy are plunging housing prices (and its knock-on effects on the financial system) and softening consumer spending. There is an easy fix that would have the additional benefits of positioning America for continued dominance of the world economy and solving the major conflicts in the world.

The fix is to change immigration policies so that:
  • Anyone from anywhere in the world who has a college degree and a net worth of at least $20,000 can get a greencard (permanent resident status) provided they emigrate to the US and buy a house in the next 6 months.
  • Anyone living in a conflict zone can get a greencard on the same conditions regardless of education or net worth, and the US will subsidize moving costs and a downpayment on a house for families below a certain net worth.
  • Special government subsidized mortgages would be available to these immigrants, as well as special loan programs for financing purchases of everything a family needs to set up a new home in the US: cars, microwaves, TVs, etc.
The surge of home buying by immigrants will stall the downward slide of the housing market, saving us hundreds of billions in bailouts and programs to save people from foreclosure.

The surge in purchases basic durable goods by the immigrants (cars, stoves, clothes, etc) will boost consumer spending right when we need it.

The immigration program for people with college degrees will provide the US with the most highly skilled workforce in the world, giving us a competitive advantage in the future (thanks to Thomas Friedman for this idea).

The immigration program for people in conflict zones will allow the people who want to escape war to escape to somewhere where they can lead a peaceful life. As the best and the brightest start to flee conflict zones, the adversaries will realize that every day of continued fighting is draining away what they are fighting over, and if they don't stop soon there will be nothing left to fight over. The conflicts that have any possible resolution will be settled quickly. The truly intractable conflicts will be concluded when the last non-combatent leaves and the combatents finish killing each other.

Will the surge of immigration increase unemployment? Sure it will in the near term. But in the long term employment will benefit because America will be the go-to country for skilled workers.

2009-02-15

What about the personal saving rate?

One thing that worries me about the current economic unpleasantness is how the personal saving rate plays into the downturn and the possibilities of recovery in the near term.

The personal savings rate is calculated by the Bureau of Economic Analysis (http://bea.gov) as personal income, less taxes, less personal outlays, divided by disposable personal income. To me, its not so much a measure of how much households save for the future.  It's more a measure of how much of each paycheck gets spent (i.e. the "average propensity to consume").  If households are, on average, living completely paycheck to paycheck, then the personal saving rate will be zero. If they spend less than they earn, then the personal saving rate will be some positive percentage.

The reason the personal saving rate is important is that it, along with personal income levels, determines the size of consumer spending (called personal consumption expenditures or PCE by the BEA), and consumer spending makes up approximately 70% of Gross Domestic Product.

Now for the scary part.

If you look at the personal saving rate for the last twenty years you will see that in 1988-1992 it was pretty solidly in the 7% range, which is pretty close to its long term average, but then starting in the mid nineties its started to drop until finally in 2005 it hit a low of 0.4%.  That means that over that period US households shifted, on average, from spending only 92.7% of each paycheck to 99.6% of each paycheck.  While on a percentage basis its just 7 percentage points, in dollars its huge: 7% of the 2008 personal disposable income equals 744.6 billion dollars, or about 5.2% of GDP.

So, what happens if US households revert to the historical average personal saving rate of around 7%?  By my amateur calculations there will be a minimum contraction in GDP of 5.2% (just subtracting the reduced consumer spending from today's GDP), before counting multiplier effects and without counting any knock-on reduction in personal disposable income from lost jobs, which means in reality the GDP drop would probably be substantially higher.

This indicates to me that if people stop living paycheck-to-paycheck we are going to have a deep contraction of GDP, and we will not return to previous GDP levels for a few years at least. And once again, that before counting feedback and multiplier effects (declining GDP and declining personal income).

Once I started worrying about the personal saving rate the questions in my mind became:
  1. Why did households start living paycheck-to-paycheck in the first place?
  2. Are households pulling back from living paycheck-to-paycheck, and if so, why?
  3. What are the odds of households going back to living paycheck-to-paycheck soon enough to avoid a long recession?
Why did households start living paycheck-to-paycheck in the first place?

My thesis is that the personal saving rate collapsed in the period 1988 to 2005 in part because US households bought in (literally) to the story of a stock market that reliably returns 10% and housing prices that reliably rise year after year.  According to this source equity ownership among US households rose from 29.6% in 1989  to 56.9% in 2005. This was primarily fueled by the explosion in participation in 401k plans.  Households saw their 401k statements, and their home equity, growing briskly over time, and were constantly being told by multiple sources that a long term 10% annual return on stock market investments was a sure thing.  In this environment of swelling 401ks and home equity values saving $7 out of every $100 seemed kind of dopey and old fashioned.  The smart thing to do was to stop saving so much of each paycheck and instead "put your home equity to work for you" by borrowing against it to finance consumption, and trust that a combination of reliable stock appreciation and home value appreciation would take care of retirement savings.  I distinctly remember reading articles a few years back about how the declining personal saving rate was nothing to worry about in terms of Americans saving for the future because the appreciation in the stock market and housing meant that Amercan's wealth was growing faster than ever despite the reduction in personal saving.

The other factor that I theorize helped the personal saving rate collapse in the period 1988 to 2005 was the explosion of ways for consumers to borrow.  One reason to not live paycheck-to-paycheck is that if your life can get ruined if you get caught out by some large unexpected expense like a car dieing, or a major repair.  However, as credit cards and home equity loans became much easier to get, the smart thing to do became to stop saving for a rainy day, and instead just draw on a home equity loan or credit card to get you through any rough patches.  There were certainly many experts advising consumers to do just that in the last few years. I theorize that essentially what happened was that many consumers because to view their home equity and the availability of easy unsecured credit as defacto cash in the bank (i.e. very liquid money available for spending on little or no notice).  Once these consumers started seeing easy credit and home equity as the equivalent of cash the need for actual cash savings vanished.  If you poke around you can find financial advice from circa 2005 essentially advocating this view, and certainly there were a raft of ads around then from home equity lenders promoting the view that home equity should be "put to work" by borrowing against it for consumption.

There is some support for this theory in a May 2006 paper from the St. Louis Fed "Cross Cultural Personal Saving Rates" where they found "The personal saving rate usually declined more in countries where access to domestic credit grew faster."

The decline of the personal saving rate is discussed in detail in "The Decline in the US Personal Saving Rate: Is it Real and is it a Puzzle" from the November/December 2007 edition of the Federal Reserve Bank of St. Louis Review. They run through all the theories for the decline.  The first theory (stock and housing appreciation) is mentioned but discounted on the basis that it is not adequately supported by empirical evidence. Oddly, my second theory is not mentioned.  The paper discusses the increase of availability of credit, but only through the mechanism of actual borrowing being used to increase consumption.  There doesn't seem to be any mention of my theory that just the availability of easy credit can inspire people to spend more of their paychecks even if they don't actually borrow anything. In the end, the authors conclude that no existing theory can explain the decline in the personal saving rate and so its a puzzle.  Maybe my second theory of easy credit as a substitute for an emergency fund is one of the missing pieces of the puzzle?

Are households pulling back from living paycheck-to-paycheck, and if so, why?

The personal savings rate has shot up to 3.6% in December 2008. And the reasons seem pretty simple: Credit is harder to get, 401k balances have been decimated, people have much less home equity, jobs are evaporating at near-record levels, and, perhaps most significantly, some of the foundational myths that supported the lifestyle of living paycheck-to-paycheck have been busted:
  • The stock market has a long term average return of 10% in the long term. Busted.
  • There has never been a national decline in housing prices so home equity will always grow. Busted.
  • Its always easy to just borrow to get through emergencies. Busted.

For all these reasons living paycheck-to-paycheck just doesn't look so smart any more.  In addition to the busted myths, there are some new factors the promote increased personal saving:
  • High rates of job losses increase the need for an emergency fund, especially in light of tightening consumer credit.
  • People who are upside down in their houses must have a substantial emergency fund to avoid complete ruin by unexpected expenses or job losses.  There are a number of reasons for this:
    • Before the real estate bubble burst a middle class person who lost their job could go on a nationwide job search, which minimized the probability of extended unemployment.  People who are upside down in their homes don't have this option (actually given the slowness of the real estate market relocation for a job is hard for everyone now).
    • If you are upside down in your house, the chances of getting credit to tide you through a rough patch are pretty slim.
What are the odds of households going back to living paycheck-to-paycheck soon enough to avoid a depression?

In my mind this is the big big big question that everything turns on. If the personal saving rate peaks at 3.6% and then reverts back to near zero in the next few months, then it seems possible for the economy to recover in the next year or two because consumer spending will recover.  If the personal saving rate keeps climbing for the next year or two, then no amount of bailouts and/or stimulus will bring GDP back to its previous level because (1) healthy credit markets don't matter if people just don't feel like borrowing to live paycheck-to-paycheck anymore, and (2) there are no plans in the works for a stimulus package big enough to replace the lost spending from a 7+% increase in the personal saving rate.  Increased personal saving will cause GDP to fall, which will reduce personal income, which will, combined with the personal saving rate, reduce consumer expenditures, which will reduce GDP, in a cycle that will continue until the personal saving rate is forced back to zero (like it was in the Great Depression) because personal income has fallen so far that on average households have to live paycheck-to-paycheck just to procure the basics.  To avoid an very long depression, the personal saving rate has to fall back to near zero before GDP has contracted substantially. Once GDP (and personal income with it) has contracted substantially it will be too late for a reduction in the personal saving rate to save the economy because the damage will already be done.

If getting the personal saving rate back to near zero in the near term is key, what are the chances of getting American consumers to go back to living paycheck-to-paycheck in the next 12-18 months? I mentioned above about how some of the myths that support living paycheck-to-paycheck have been busted: The myth of the perpetually growing stock and housing values and the myth of perpetually available easy credit.  As long as the average household no longer believes in these myths its going to be really hard to get them to spend every penny of every paycheck, especially the baby boomers.  There are a lot of baby boomers out there in their mid to late fifties who had been living paycheck-to-paycheck who have just seen their 401k plans lose 30-40% of their value, and much, if not all, of their home equity evaporate.  And they are all reading about record unemployment levels.  If you were in their shoes would you continue to live paycheck-to-paycheck, with retirement maybe 8 years away? Some will, undoubtedly, but its the average that is important, and its hard to imagine that on average the baby boomers will continue living paycheck-to-paycheck in the coming months.

To me the important question is just how busted are these myths?  At the moment I suspect the reputation of these myths has only been damaged, but not completely destroyed, for many people.  There are probably many people who expect that we will soon return to "normal."  However, with each passing month without a recovery the reputation of these myths will continue to dim.  Although the media doesn't yet mention it often, the stock and real estate markets in Japan never did recover after their real estate bubble popped in 1990 (The Japanese stock market was still well below its peak 17 years later in 2007, and Japanese real estate prices declined for 15 consecutive years). As time passes without a recovery, the Japanese lost decade experience may tend to be seen as more and more relevant and telling, and the cultural consensus may shift more and more against the myths required to support living paycheck to paycheck. Some support for this theory that the personal saving rate tends to rise after a real estate bubble bursts can be found in the Japanese "Lost Decade."  According to a May 2006 paper from the St. Louis Fed "Cross Cultural Personal Saving Rates" the Japanese personal saving rate jumped after their 1990 crash, and then slowly climbed for almost 9 years.