Results for “Tests”
771 found

The elasticity of natural disaster deaths with respect to income

Matt Kahn has a good paper (and here) on this topic:

Using a new data set on annual deaths from disasters in 57 nations from 1980 to 2002, this paper tests several hypotheses concerning natural disaster mitigation. While richer nations do not experience fewer natural disaster events than poorer nations, richer nations do suffer less death from disaster. Economic development provides implicit insurance against nature’s shocks. Democracies and nations with higher quality institutions suffer less death from natural disaster. The results are relevant for judging the incidence of a Global Warming induced increase in the count of natural disaster shocks.

Sentences to ponder

The topic is consumer protests over price hikes for eBooks and here is one response:

“The sense of entitlement of the American consumer is absolutely astonishing,” said Douglas Preston, whose novel “Impact” reached as high as No. 4 on The New York Times’s hardcover fiction best-seller list earlier this month. “It’s the Wal-Mart mentality, which in my view is very unhealthy for our country. It’s this notion of not wanting to pay the real price of something.”

Here is the article; I giggled when I read that.  Here is a short biography of Douglas Preston.

The AP critique of the stimulus

There is apparently a new study, from the AP, suggesting that the transportation spending of the stimulus has not succeeded in creating jobs.  The study now seems to have "legs," as here is lengthy NYT coverage.  Through some email forwarded to me, I have the impression (my apologies if I am wrong) that they are not circulating copies of the study for perusal.  Instead, the study has been reviewed by some economists, who seem to approve of it.  No one else is allowed to judge.  Does anyone have a copy of the original study?  Will the AP send me a copy?  The NYT piece — which seems to be written by the AP by the way — does not link to the study.  The AP won't link to their own study, or so at least it seems.

Loyal MR readers will know that I have been critical of most of the stimulus program.  Still, phantom studies should not be receiving serious media attention.  It's time for the AP to put up or shut up.

Comments are open, as is my email, I would like to see a copy of the study.

Addendum: from Matt Apuzzo, at the AP:

"Matthew/Tyler:

If either of you would like to chat about the AP's analysis, I'm happy
to provide you the sources of our data and walk you through the
statistical tests we conducted. Nothing we did is a secret, but there's
no actual "study" to provide you, like there is in academia. The
professors reviewed some spreadsheets and statistical tests, talked
methodology, made suggestions on other tests to run, and overall made
sure we weren't reading things incorrectly. All of this feedback
ultimately contributed to our final conclusion, but there's no executive
summary.

Give a call, I'm happy to help.

-Matt"

What do you all think?

Markets in everything China fact of the day

Wanted: One live-in protester, $146 a month, no days off.

When the managers of a Beijing restaurant marked for demolition were too busy to fight it, they posted an Internet ad and hired a stranger to stay there around the clock. The job seems to be a first for China, where frenzied urban construction has led to violent evictions, protests and even suicide.

Huddled on a makeshift bed in the trash-strewn, freezing restaurant, Lu Daren said he once worked for a demolition crew and understands their tactics.

"I'm tired," the 46-year-old said Thursday, after a long night of fending off the latest visit from what he suspects were hired thugs by the landlord. "Tired, tired, tired." He stays – wrapped in blankets, reading the newspaper or writing idle poetry, occasionally taking short walks_ because he thinks the restaurateurs have been treated unfairly.

The full story is here and I thank Daniel Lippman for the pointer.

How worried should we be about the deficit?

There have been many posts on this topic lately, start with Paul Krugman and Brad DeLong if you need to catch up.  Today I have a few simple points:

1. Even if "it is fine to borrow more" is the most likely scenario, it is not the only scenario.  Let's take a page from Marty Weitzman on climate change.  The worst-case scenarios matter too, because they can be very, very bad.  We need to think probabilistically about this issue.

2. Are there current intelligent discussions of the implied interest rate volatility embedded in current options prices?  If we are looking for market tests, why not start there?  Focusing on the point estimate of the market interest rate(s) discourages you from thinking probabilistically.

3. I know less about Belgium but I am not reassured by Krugman's point that "Italy can do it."  I and many other observers consider Italy's economy to be a basket case which will only get worse.  Nor is Japan in a satisfactory place, economically speaking.

4. Krugman writes: "Belgium is politically weak because of the linguistic divide; Italy is
politically weak because it’s Italy. If these countries can run up
debts of more than 100 percent of GDP without being destroyed by bond
vigilantes, so can we."

I would interpret this evidence differently.  A high deficit often is an unfavorable symptom of bad politics, even if you think the high deficit is economically OK on its own terms.  It's a sign that you have dysfunctional institutions and decision-making procedures, as indeed they do in Belgium and Italy.  I believe that the not-always-swift American voter in fact understands high deficits — correctly — in this light.  They don't hold theories about "crowding out," rather they sense something in the house must be rotten.  And so they rail against deficits, as do some of their elected representatives.  It's a more justified reaction than the pure economics alone can illuminate.

When water regularly overflows from your toilet, you want the toilet fixed, whether or not the water is doing harm.

Caplan on Education

How much does increasing college-going rates matter to our economy and society?

Caplan: College attendance, in my view, is usually a drain on our economy and society. Encouraging talented people to spend many years in wasteful status contests deprives the economy of millions of man-years of output. If this were really an "investment," of course, it might be worth it. But I see little connection between the skills that students acquire in college and the skills they'll need later in life.

Much more here, including answers from Charles Murray, Richard Vedder and others. Hat tip to Arnold Kling.

Yglesias channels his inner Robin Hanson

Matt Yglesias offers wisdom on cutting health care spending.

Still, though waste is a huge element of our insurance spending, insurance-related waste is a relatively small portion of the overall waste–about 14 percent. The biggest chunk of excess spending we’re involved with is spending on “outpatient care.” We pay doctors more than other people do, our doctors order more tests than other doctors do, our tests are more expensive than other people’s tests, and we have many more relatively expensive specialists and relatively few relatively cheap GPs. And we have nothing to show for it.

The prospects for changing this, however, don’t look great to me. People don’t like insurance companies. Taking them on is popular. And nevertheless we see how difficult it is to really hurt their interests. Now imagine taking on the doctor lobby. More money is at stake. And doctors have a much better public image. And doctors and there families are a much bigger voting block than insurance executives and their families. And on top of that, people have a very strong mistaken intuition that getting lots of tests and seeing lots of specialists is in their interests.

Teacher Performance Pay: Experimental Evidence from India

In an impressive new paper, Karthik Muralidharan and Venkatesh Sundararaman provide evidence on the power of teacher incentives to increase learning.  The paper is impressive for three reasons:

1) Evidence comes from a very large sample, 500 schools covering approximately 55,000 students, and treatment regimes and controls are randomly assigned to schools in a careful, stratified design. 

2) An individual-incentive plan and a group-incentive plan are compared to a control group and to two types of unconditional extra-spending treatments (a block grant and hiring an extra teacher).  Thus the authors can test not only whether an incentive plan works relative to no plan but also whether an incentive plan works relative to spending a similar amount of money on "improving schools."

3)  The authors understand incentive design and they test for whether their incentive plan reduces learning on non-performance pay margins.

The results are as follows:

We find that the teacher performance pay program was highly effective in improving student
learning. At the end of two years of the program, students in incentive schools performed
significantly better than those in comparison schools by 0.28 and 0.16 standard deviations (SD)
in math and language tests respectively….

We find no evidence of any adverse consequences as a result of the incentive programs.
Incentive schools do significantly better on both mechanical components of the test (designed to
reflect rote learning) and conceptual components of the test (designed to capture deeper
understanding of the material),suggesting that the gains in test scores represent an actual
increase in learning outcomes. Students in incentive schools do significantly better not only in
math and language (for which there were incentives), but also in science and social studies (for
which there were no incentives), suggesting positive spillover effects….

School-level group incentives and teacher-level individual incentives perform equally well in
the first year of the program, but the individual incentive schools significantly outperformed the
group incentive schools in the second year….

We find that performance-based bonus payments to teachers were a significantly more cost
effective way of increasing student test scores compared to spending a similar amount of money
unconditionally on additional schooling inputs.

Surprisingly, since absent teachers are a big problem in India, reduced teacher absenteeism per se does not appear to be the primary mechanism by which incentives improve learning.  Instead the primary mechanism appears to be more intensive teaching, including more homework and classwork and better attention to weaker students, this greatly increases the relevance of these results to teaching in the developed world.

Addendum: See also Karthik's comments on the comments at 26.

Did the structure of banker pay cause the crisis?

This is an old topic but it is in the headlines again, so I pass this along, from Jeff Friedman:

This “executive compensation” theory of the crisis
is now the keystone of the conventional wisdom, having been embraced by
President Obama, the leaders of France and Germany, and virtually the
entire financial press. But if anyone has evidence for the
executive-compensation thesis, they have yet to produce it. It’s a
great theory. It “makes sense”–we all know how greedy bankers are! But
is it true?

The evidence that has been produced suggests that it is false.

For
one thing, bankers were often compensated in stock as well as with
bonuses, and the value of this stock was wiped out because of the
investments in question. Richard Fuld of Lehman Brothers lost $1
billion this way; Sanford Weill of Citigroup lost half that amount. A
study by René Stulz and Rüdiger Fahlenbrach[3] showed that banks with
CEOs who held a lot of stock in the bank did worse
than banks with CEOs who held less stock, suggesting that the bankers
were simply ignorant of the risks their institutions were taking.
Journalists’ and insiders’ books about individual banks[4] bear out
this hypothesis: At Bear Stearns and Lehman Brothers, for example, the
decision makers did not recognize the risks until it was too late,
despite their personal investments in the banks’ stock.

The Stulz and Fahlenbrach abstract reads as follows:

We
investigate whether bank performance during the credit crisis of 2008
is related to CEO incentives and share ownership before the crisis and
whether CEOs reduced their equity stakes in their banks in anticipation
of the crisis. There is no evidence that banks with CEOs whose
incentives were better aligned with the interests of their shareholders
performed better during the crisis and some evidence that these banks
actually performed worse both in terms of stock returns and in terms of
accounting return on equity. Further, option compensation did not have
an adverse impact on bank performance during the crisis. Bank CEOs did
not reduce their holdings of shares in anticipation of the crisis or
during the crisis; further, there is no evidence that they hedged their
equity exposure. Consequently, they suffered extremely large wealth
losses as a result of the crisis.

It's entirely fair to argue that these tests are not decisive.  But still, the evidence isn't there — at least not yet — that executive pay was in fact the big problem.

I thank Jeff Friedman for the pointer.

The public health insurance plan

Here is comment from Ezra Klein, who distinguishes different versions of the public plan idea and also links to further reading.  Matt Yglesias comments in favor of the idea.  Here is a Paul Krugman column.  Arnold Kling is skeptical.  Those are good introductions to the debate.  On the economics, Ezra writes:

Rather, the theory here is simple: If you can't replace them, convert
them. If the public plan works, then private insurance will work better
as well. In this telling, the simple existence of the public plan
forces a more honest insurance market: Private insurers need to offer
premiums closer to their marginal cost, and they have to cut
administrative costs, and they have to work on their reputation for
cruelty and capriciousness. The existence of another option changes the
market. Individuals will have access to private insurers, but they'll
no longer be stuck with them.

I believe Ezra is assuming no direct cash subsidy to the public plan, lower marginal and average costs for the public plan, and some mix of market power and X-inefficiency in the private insurance companies.  The existence of the public plan then "contests the market," which eventually lowers MC in the private plan and leads to lower prices and better service.

My question is what the equilibrium looks like.  Say the public plan has a cost advantage (both MC and AC), as plan proponents suggest.  If public and private plans are to coexist, the public plan must be attracting the higher-cost customers, namely the higher medical risks.  (I am also assuming that the political equilibrium does not allow the public plan to reject these customers outright.)  There is then market segmentation and it is not obvious that there are significant positive competitive pressures on private insurance companies.

Oddly, I believe in some models the public insurer constrains the private companies more tightly when the public insurer does not have an apparent cost advantage.  Even here, the properties of the monopolistically competitive equilibrium would be very tricky.

You might wonder why the public plan does not attract all the low-risk customers and take over the whole market.  I would say that either a) it does, or b) it is tailored toward the high-risk customers.  Since public plan advocates sincerely and correctly claim the policy is not just a back door to single-payer, we are left with b).

Another question: is the "cruelty and capriciousness" of the private plans — cited by Ezra — driven by profit maximization?  Presumably it is and again assume the government plan will not do the same.  Why then would public sector competition force a private firm to throw out a profit-maximizing strategy?  In fact "cruelty and capriciousness" would be a comparative advantage of the private companies and maybe it would be milked more strongly in a more competitive environment. 

Another possibility is that the public company has a bigger cost advantage on AC than MC.  For instance maybe it has a "head start" on the fixed costs, because everyone has heard of it, but its cost advantage for additional service dwindles at some point.  The successive accretion of high-risk customers then threatens to put the public plan under (especially if there are lots of previously uninsured and they are high risks) and the public plan requires a subsidy simply to break even.  I consider this equilibrium to be not totally unlikely.

Obviously I am missing some equilibria, but in many cases the public plan is mainly providing insurance to high-risk customers.  There's nothing wrong with that (and indeed it is a major policy goal), but the resulting equilibrium needn't much improve the performance of private health insurance.  I file this argument under "not yet established."

Fearless Critic

The subtitle is Washington DC Area Restaurant Guide and the author is Robin Goldstein.  I am a Contributing Editor and yes he did listen to my most valuable pieces of advice.  Described as "brutally honest," this is much, much better than Zagat's and the like.  It is the best book of its kind.

Elsewhere on the new book front, there is Keith Stanovich's What Intelligence Tests Miss (I hope to review it) and Robert Wright's The Evolution of God; there is some chance I will be doing a BloggingHeads with Wright on this book.

Divorce and Crime Victimization

While paging through the statistical tables of Criminal Victimization in the United States I found some interesting data on victimization, marriage and divorce.  The rate of victimization for violent crimes (per 1,000 persons aged 12 and over) for never married and married males is as follows:

  • Never Married Males: 45.0
  • Married Males: 12.3
Clearly, married males are older and they have settled down, usually in places away from crime hot spots.  Thus the fact that the rate of victimization for married males is much lower than for never married males is no surprise.  What did surprise me is that divorced males have rates of victimization about as high as for never married males:
  • Divorced or Separated Males: 44.2
The same pattern is even stronger for females:
  • Never Married Females: 38.4
  • Married Females: 10.3
  • Divorced or Separated Females: 49.4

The patterns are suggestive of how large a difference one’s choices can make for criminal victimization.  That is, one hypothesis to explain the data is that singles congregate in urban, high crime areas and they go out at night to bars and other high crime locations.  Married individuals move to low crime suburbs and stay home with popcorn and Netflix.  The divorced, however, move back to the cities where the singles are and they head out at night to try to mate again.

An alternative hypothesis is that the individuals who tend to get divorced have personalities or behaviors which make them more likely to get divorced and more likely to be victims of crime: a drug user, for example, is likely to have a higher probability of divorce and a higher probability of being a victim of crime than a non drug-user.

How many other hypotheses can you think of to explain the data?  What tests would you suggest to distinguish hypotheses?

The relative value of health care

In a 2003 study, another Dartmouth team, led by the internist
Elliott Fisher, examined the treatment received by a million elderly
Americans diagnosed with colon or rectal cancer, a hip fracture, or a
heart attack. They found that patients in higher-spending regions
received sixty per cent more care than elsewhere. They got more
frequent tests and procedures, more visits with specialists, and more
frequent admission to hospitals. Yet they did no better than other
patients, whether this was measured in terms of survival, their ability
to function, or satisfaction with the care they received. If anything,
they seemed to do worse.

That’s because nothing in medicine is
without risks. Complications can arise from hospital stays,
medications, procedures, and tests, and when these things are of
marginal value the harm can be greater than the benefits. In recent
years, we doctors have markedly increased the number of operations we
do, for instance. In 2006, doctors performed at least sixty million
surgical procedures, one for every five Americans. No other country
does anything like as many operations on its citizens. Are we better
off for it? No one knows for sure, but it seems highly unlikely. After
all, some hundred thousand people die each year from complications of
surgery—far more than die in car crashes.

To make matters worse,
Fisher found that patients in high-cost areas were actually less likely
to receive low-cost preventive services, such as flu and pneumonia
vaccines, faced longer waits at doctor and emergency-room visits, and
were less likely to have a primary-care physician. They got more of the
stuff that cost more, but not more of what they needed.