Depressing education chart of the day

Mike the Mad Biologist points us to this one. 

From Karl Alexander (John Dewey Professor of Sociology at Johns Hopkins University):

We discovered that about two-thirds of the ninth-grade academic achievement gap between disadvantaged youngsters and their more advantaged peers can be explained by what happens over the summer during the elementary school years.

I also want to point out that the higher performing group isn’t necessarily high income, but simply better off. In the context of the Baltimore City school system, that usually means solidly middle class, with parents who are likely to have gone to college versus dropping out.

Statistically, lower income children begin school with lower achievement scores, but during the school year, they progress at about the same rate as their peers. Over the summer, it’s a dramatically different story. During the summer months, disadvantaged children tread water at best or even fall behind. It’s what we call “summer slide” or “summer setback.” But better off children build their skills steadily over the summer months. The pattern was definite and dramatic. It was quite a revelation.

Mike goes on to point out the obvious question this raises about the fire-the-teachers approach to education reform.

UPDATE: This seems like a good time to remind everyone about some of the other previously mentioned challenges these kids face:

Hart and Risley also found that, in the first four years after birth, the average child from a professional family receives 560,000 more instances of encouraging feedback than discouraging feedback; a working- class child receives merely 100,000 more encouragements than discouragements; a welfare child receives 125,000 more discouragements than encouragements.

More on Hostess

There is a rumor that Mark is preparing a nice post on the whole Hostess bankruptcy story.  So, to continue to generate context, consider this story from an actual employee:

In 2005 it was another contract year and this time there was no way out of concessions. The Union negotiated a deal that would save the company $150 million a year in labor. It was a tough internal battle to get people to vote for it. We turned it down twice. Finally the Union told us it was in our best interest and something had to give. So many of us, including myself, changed our votes and took the offer. Remember that next time you see CEO Rayburn on tv stating that we haven't sacrificed for this company. The company then emerged from bankruptcy. In 2005 before concessions I made $48,000, last year I made $34,000. My pay changed dramatically but at least I was still contributing to my self-funded pension.

In July of 2011 we received a letter from the company. It said that the $3+ per hour that we as a Union contribute to the pension was going to be 'borrowed' by the company until they could be profitable again. Then they would pay it all back. The Union was notified of this the same time and method as the individual members. No contact from the company to the Union on a national level.

This money will never be paid back. The company filed for bankruptcy and the judge ruled that the $3+ per hour was a debt the company couldn't repay. The Union continued to work despite this theft of our self-funded pension contributions for over a year. I consider this money stolen. No other word in the English language describes what they have done to this money.
 

 This illustrates a couple of things.  One is that defined contribution pensions (i.e. the 401(k) and other such vehicles) are not necessarily a safe harbor from corporate games if something like this can happen without the union proactively agreeing.  Two, the concessions made by the unions were pretty significant.  They took pretty large pay cuts (29%) and lost additional money from the pension fund.  No manager who made the decision to borrow this money was held liable. 

Maybe the company was in an impossible situation.  That is definitely one interpretation of the situation but the narrative is odd for this outcome.  The argument that unions were to blame is hardly the case unless this account (above) is wracked with falsehoods.  If the situation was hopeless then why did the high paid executives not call it over before raiding the pension?  Was it because a longer tenure of salaries and job experience would be of personal benefit to them and they were no liable for the consequences of their decisions? 

There are no easy answers, but a simple narrative of union greed or dysfunctional bargaining is clearly omitting lots of nuance.  If nothing else, reneging on the repayment of the pension fund borrowing was hardly the ideal prelude to asking the union to trust future management promises. 

Hoisted from the comments

A rather nice point from commenter Bill Trudo that I thought was worth emphasizing:

The problems at Hostess have been decades in the making. The company went into bankruptcy in 2004 and left in 2009 with numerous labor concessions, but without major reform to the pension system. Some people liked to blame that on the unions, but it is up to management to decide what is profitable or unprofitable. Now three years later, management has deemed their old agreements aren't profitable and that would be correct given how much money Hostess has been bleeding. And people still wonder why some union leaders were skeptical of management.
Skepticism is warranted, especially in light of Hostess having left bankruptcy in 2009 with more debt than it had in 2004 with falling sales. 2011 sales were down 28% from 2004. Of course, this is the union's fault that sales have dropped and not a failure of the company's leadership.
On top of that, management comes to the labor unions demanding 8% pay cuts and slashes to the pension, while the top executives increase their own pay--what a nice goodwill gesture. Some unsecured creditors actually filed suit over the pay raises.
Yes, pensions and work issues needed to be addressed. The unions had to make concessions, but it's difficult in light of what has happened for anyone to have any faith in what management has been doing at Hostess.
Now, enter two hedge funds, who hold the majority of secured debt, and they're likely looking to get out. After all, the business model isn't working; it's time for a showdown.
The Teamsters actually finally agreed to the wage and benefit concessions. Some of the other unions, including the Bakers' Union, didn't. Of course, I have a suspicion that some of the unions were acting in the best interest of all their union workers and not the union workers who were working at Hostess. Though, I also have suspicions that even if the unions capitulated to the hedge fund demands (they're the ones driving the show), Hostess would have likely been sold or have been back in bankruptcy in a few years.

This really gets to the core of the union/management dilemma -- it is hard to take things on faith.  I remember many of my co-workers (including myself) putting in huge amounts of work for promises that never materialized.  The classic "if you take a hit now we will remember you once this bad period is over".  Curiously, it almost never works that way.

I also note that the narrative is "bad union wouldn't accept huge cuts" and not incompetent management was unable to make a top brand into a successful business.  This is especially true of pension cuts.  Underfunded pensions happen for a lot of reasons but they are usually much more of a management decision than a union decision.  Maybe it was a bad decision for management to agree to the pension plan in the first place -- that is certainly possible.  But it is always concerning to see no ethical qualms about removing obligations to workers (breaking agreements for work already done) while increasing your own wages.  

What's with the (non)weird name?

Hi,

The regulars may have noticed something different about the top of the page. Observational Epidemiology is now West Coast Stat Views. This doesn't represent a change in content, rather a nod to truth in advertising. When this blog started, about two thousand posts ago, the name was both accurate and descriptive,   but it soon morphed into a general statistics blog with a focus on topics that interested both of us and which connected (at least indirectly) with some relevant experience of one or both of us.

Both Joseph and I started out as corporate statisticians, so not surprisingly we're both mildly obsessed with business and economics. We both spent a great deal of time in academia so we're interested in education. As you'd expect from a blog originally called Observational Epidemiology, health care remains a big issue. As for politics and journalism, everybody needs a hobby.

Besides, I was getting tired of explaining what observational epidemiology meant.

Happy Holidays,
Mark

"Peggy" "Noonan's" Life on Mars

Apparently Martians are also using the analytic and marketing methods we've been using since the late 20th Century.

From the Wall Street Journal (via DeLong via Kos)

The other day a Republican political veteran forwarded me a hiring notice from the Obama 2012 campaign. It read like politics as done by Martians. The "Analytics Department" is looking for "predictive Modeling/Data Mining" specialists to join the campaign's "multi-disciplinary team of statisticians," which will use "predictive modeling" to anticipate the behavior of the electorate.

It might seem that, after the last thousand or so Noonan columns, pointing out flaws in the latest entry is coals-to-Newcastle. This is, after all, the pundit who told us the day before the election that Romney was on track to win because "all the vibrations are right."  (She said a lot of other memorable things in that column. You really ought to check it out.)

The temptation is to start letting them slide, but there's an issue here of reputation and metadata. Noonan is recognized as a authority on politics and her reputation is backed by the reputations of the major news outlets that publish her writing and put her in front of the camera. All of this contributes to the metadata that tells us how much weight to give Noonan's statements.

For the system to work, information sources that are consistently unreliable have to take a reputational hit, but Noonan either knows nothing about how politics works today (which is, remember, her area of expertise) or, worse yet, she has such a low opinion of her readers' intelligence that she believes she can scare them by dropping a few technical terms.

If Noonan's reputation doesn't take the hit here then the reputations of those who publish and broadcast her will. Every time the Wall Street Journal prints one of these columns, the trust we can place in the paper diminishes just a little bit. It's a small effect but it's cumulative. Give it long enough and it can eat away the standing of even the Journal.

Go read Noah Smith

Today's post is really good:

Sicko is about America's health care system, and the alternatives. Before I saw Sicko, I believed the common line that, for all its flaws, America's health care system was "the best in the world". After I saw the movie, I did not believe anything of the kind. Sicko opened my eyes to the existence of Britain's National Health Service; after watching the movie, I looked into the NHS, and found that it achieves better results than the U.S. on almost any outcome measure, for far fewer costs. Importantly, it does this using a rational incentive system - doctors are paid for improving the health of their patients, not for recommending large numbers of expensive services.
I'm not sure, but around the same time that I stopped believing that America had the best health system in the world, I noticed that other people stopped saying it (and in fact started saying the opposite!). Around the same time I started thinking that Britain's NHS is the best alternative, I noticed a lot of policy-wonkish people praising that system in the press. Around the same time I started realizing the insanity of the "fee for service" incentive system, everyone started talking about it. So I wonder if Sicko, rather than just changing my mind, actually changed the whole national conversation

This really highlights the central dilemma facing US medical care.  It is really expensive and it has very mediocre outcomes at a population level.  It is possible that some people get exceptional medical care beyond that available elsewhere.  And it is true that putting a lot of resources into a sector does tend to drive advances in that sector.  My own person question is whether this is the best way to drive medical innovation; could targeted research money do more good on innovation than paying more for physician office visits?

Once again I notice that Canada has very tough drug patent protections (which might do a lot to improve innovation in the drug arena) and still manages to have much lower costs for medical care (as a percentage of GDP).  Canadian Healthcare has some major downsides and I would never endorse it as the best possible system (I actually prefer the British NHS for a lot of reasons, that I really need to blog about).  

Jon Stewart opens up a big ol' can of historical context on Bill O'Reilly

An absolutely beautiful piece of satiric commentary.

The Daily Show
Get More: Daily Show Full Episodes,Political Humor & Satire Blog,The Daily Show on Facebook

Follow-up to yesterday's post

The discovery that Hostess tripled CEO pay (and boosted the pay of other senior executives) while blaming overpaid unions for the end of the company is a great example for yesterday's point.  It would be fine to blame everyone involved.  But some of the issues, like under-funded pension obligations, strike me as more of a management issue than a worker issue.  But, regardless, it is striking that the narrative is all about the striking bakers and not the huge cash grab at the executive level.

EDIT: See Thoreau as well.

Free Market Wages

I often get frustrated by defenses of extremely high pay as needed to be able to attract the incredibly small pool of high skill workers.  This comment is pitch-perfect:

The claim that high pay is necessary to attract and motivate good workers is not merely an economic postulate. If it were, it could apply across the wage distribution. Instead, it functions as a defence of inequality - used to justify higher pay for the rich, rather than for any job. For grunt workers, wages are a cost to be minimized regardless of consequence.

This is a really good point; if wages for managers were treated like an expense for the company (instead of as an investment in talent) then would we have seen manger salaries rise so far in recent years?

Why do we not have the same sort of bidding war for mid-level talent?  I have the same question in a different way: why are managers getting high wages is seen as a sign of paying for talent but union workers are seen as using sleazy tactics to work the system?  Do we see corporate boards setting salaries for CEOs as being fundamentally different than governments setting wages for government workers?  In both cases a group of experts are setting wages based on expert opinion and not on free market principles.

Today's tuition discussion

From Counterparties:

Megan McArdle argues that surging costs are as much a consequence as they are a cause of unprecedented levels of student debt, spurred on by government subsidies. But Mike Konczal flags a Department of Education study that shows that the government earns $1.14 back for every dollar it loans to students and asks, “What’s a good word for the opposite of a subsidy? Whatever it is, student loans are that”.
 

 This really seems to be a major problem with the modern discourse on education.  We are all convinced that there has to be some sort of amazing (clever, counter-intuitive) theory about why tuition is going up.  Bad policy, on the other hand, seems to be completely ignored.  But if the government is making a net profit off of student loans, that seems to be rather concerning.  Not because I object to profit, but because the loans are so high.

There was a period where this blog used to use the word "Canada" a lot.  The reason is that Canada has a lot of interesting existence proofs for US policy.  They have universal health care, acceptable health outcomes, and a much lower percentage of GDP spent on health care.  Their system has flaws and trade-offs, but it shows that a diverse, multi-ethnic and geographically large country could control health care costs. 

In the same sense they seem to have much lower tuition and very high quality Universities.  They may not be the absolute best, but they are surprisingly competetive for a small country.  I am not sure we'd want to adopt all of the problems of Canadian education, but it is worth noting that they seem to be able to deliver high quality education at a much lower price point.  And this should have alerted the clever reader to the problem of Ms McArdle's argument -- why doesn't it happen in other countries, which also have a high return to education? 

Social Security: really, what is going on here?

I keep asking the same question as the commenter on The Incidental Economist asks:

In fact, Social Security is NOT in crisis. If we hiked payroll taxes by two percentage points we would probably take care of the payroll tax deficit. Yes, that’s a regressive tax and we should be cautious about doing it. But our long-run budget problems are mostly in health care. Why the unearthly fascination with cutting Social Security?
 

I mean, seriously, why is it such conventional wisdom that it needs to be reformed?  It is true that a private alternative would be a great revenue center for finance types.  But a lot of things would be great revenue sources that we don't privatize for good reason (Police, Military), either because it would be inefficient or because only the government can guarentee long term contracts like those required for retirement in 45 years. 

Social security is a popular benefit and something like 80-90% of the population will eventually have it as an important piece of their income.  Just what is going on here? 

The biggest political story you probably haven't heard

(At least if you're not on Pacific Time) were California propositions 30, 32, and 36. I don't have time to give this the write-up it deserves, but the LA Times voting guide is a good starting point. The short version is this: under long odds  (including the worst piece of vanity politics since Nader hung it up and a special cameo appearance from the Koch brothers) and by significant margins, the largest state in the union has just made a sharp turn to the left on taxes, education, unions and mass incarceration.

What's more, I suspect that the interest in these issues may have pumped up voter turn out and, as a result, helped Obama's popular vote totals. Add to that a potential supermajority in both houses.

None of this may be as interesting as genetically modified food and condoms for porn stars but it's a damned sight more important.

You really should be reading Andrew Gelman's blog

I considered this point (made by Andrew Gelman) to be really interesting:

All statisticians use prior information in their statistical analysis. Non-Bayesians express their prior information not through a probability distribution on parameters but rather through their choice of methods. I think this non-Bayesian attitude is too restrictive, but in this case a small amount of reflection would reveal the inappropriateness of this procedure for this example.

I had never phrased things like this but it does seem to be a sensible description of what I (essentially a Frequentist) actually do in real life.  This may be a very good teaching point and it nicely illustrates that all analysis involves priors -- just that some are more explicit than others.

More for blogger's to-do list

Really swamped this week, which invariably means that all sorts of interesting topics have been popping up. Here's a partial list. I'll try to get to as many as I can, but if any fellow bloggers want to beat me to the punch please go for it.

1. The ethics (and advisability) of psychoanalyzing your opponents
(when is it acceptable, wise and not overly sleazy to try to explain your opponents' actions using psychology and sociology and how do you avoid using those explanations to dismiss valid arguments on the other side)

Link to Pauline Kael's "Raising Kane" (no, really)


2. Implications of a submerging Republican majority
A. Scarcity and cognitive dissonance -- link to Cialdini
B. The looting phase (or Karl Rove is not your friend) -- link to Red State, Rick Perlstein, David Frum
C. Waiting for the Big Dog  -- link to Ed Kilgore (at some length)
D. Nate Silver and the Sandy Hypothesis  -- link to Talking Points and 538

3. The killer app for the driverless car -- it's not what you think

4. Daniel Engber doesn't seem to get the subtleties of statistics debates -- link to this and this

5. Nudge cars (I'll explain later)

6. Homemade Pi -- a classroom exercise for geometry, spreadsheets and Monte Carlo techniques

7. More Groupon -- link to Felix Salmon

8. Trivial thoughts -- athletic actors

9. The annual Toys-for-Tots post

Podhoretz mans up

Lots of pundits did shoddy work covering the past election. Very few have actually owned up to it.

From NPR:

FOLKENFLIK: But if Florida stays blue, Silver will have picked every state correctly, along with the president's margin of victory. Conservative columnist John Podhoretz of the New York Post and Commentary magazine had earlier argued pollsters were getting it wrong by ignoring the high turnout by Republicans in the 2010 elections that swept the GOP into control of the U.S. House. The 2012 race would be the same, Podhoretz argued, quite mistakenly, as it turned out.

JOHN PODHORETZ: That view was strengthened and amplified by what I wanted to happen, which I freely confess. People don't ordinarily cast a skeptical eye on data and information that supports their opinions. They're happy to take it.

p.s.  Add Unskewed Polls' Dean Chambers to this list.:

“Most of the polls I ‘unskewed’ were based on samples that generally included about five or six or seven percent more Democrats than Republicans, and I doubted and questioned the results of those polls, and then ‘unskewed’ them based on my belief that a nearly equal percentage of Democrats and Republicans would turn out in the actual election this year,” Chambers wrote on The Examiner website. “I was wrong on that assumption and those who predicted a turnout model of five or six percent in favor of Democrats were right. Likewise, the polling numbers they produced going on that assumption turned out to be right and my ‘unskewed’ numbers were off the mark.”