Wine, Physics, and Song

Via Kyle Deas’ Fitful Murmurs, a link to an article in the Financial Times reporting on a study of celebrity “naked chef” Jamie Oliver’s “Feed Me Better” pilot program of serving fresh-cooked, nutritious lunches in British schools.  (The FT article reports that it involved a special “boot camp” training session for the dinner ladies.)  I think I first heard about this program, and perhaps the research study, from my wife.  There’s a strong suggestion that it yields a few percentage points improvement in the number of students reaching particular levels in Science and English testing, and a more detailed examination of the regressions in the original article looks broadly consistent with that (even the “non-statistically significant” coefficients, including ones in Maths, were almost all in the right direction and mostly of about the same magnitude).  The study compared the two-academic-year periods of 2002-2004, to 2005-2007, in five school districts (”local educational areas”, LEAs) in the southeast of London—Greenwich, in which Oliver’s fresh-cooked meal program was implemented beginning in the 2004-2005 school year, with four others in which it wasn’t.  The change in absenteeism in Greenwich, relative to the change in other districts, was estimated to be  -1.2 percentage points, with a standard deviation of estimate of 0.365%, significant at p<0.01.  This is a pretty significant drop–not a 1.2 percent drop in amount of absenteeism, but a change of 1.2 percentage points in the absenteeism rate—translating to about a 22.5% drop in the number of absentee students.   (In the text, the authors cite only the 0.8 percent drop in authorized absences, since the change in unauthorized ones wasn’t significant on its own; but their data shows the total change as significant too, which is what I described.) Even more interestingly,  the estimated change, relative to non-treated schools, in the fractions of students reaching Levels 4 and 5 on the “Key Stage 2″ standardized testing, broken down into English, Maths, and Science (for four coefficients) ranged from 2-6%, with standard deviations of these estimates around 2.5-3%, in a “school level data” regression analysis.  (This means other factors that might have affected results were included in the analysis by including variables for school characteristics in the regression.)  Only a couple of these six coefficients were significant (at the 0.10 level), but I think the overall pattern for the fractions of students attaining Level 4 and Level 5, with those coefficients that were not significant at 0.10 still large and in the right direction and roughly comparable to the standard deviation of estimate, adds persuasiveness to the results.  Level 3 fraction was also included, but the effects were negligible compared to the standard deviation of estimates.   (I haven’t looked at the methodology carefully enough to determine whether Level 3 meant “exactly level 3″ or “at least level 3… in the former case, the rough constancy at level 3 is perhaps good news as it would suggest students who scored at level 4 or higher rather than level 3, were at least roughly counterbalanced by ones who reached 3 rather than a lower level.)  Pupil-level regressions were also done, with the pupil’s percentile score as dependent variable, and pupil-level data to help control for socioeconomic status and other confounding variables; it gave broadly similar results:  the relative effect on English results was large at 4.7 percent, and highly significant at p<0.01, while the Science doesn’t look too bad either at 3.6 percent (apparently not significant even at 0.1 though) and even the maths is in the right direction.

I would have liked to see more explicit statements on the school and pupil-level characteristics included in the regressions, but I take the school-level characteristics to have been those summarized in Table 2.  Actually, if we interpret “controls include” to mean that this is the full list of controls, then the information is in the note at the bottom of Table 3.

I’m not a practicing statistician or econometrician, so I won’t pretend to give a definitive assessment of the methodology here (I also haven’t spent that long thinking about it), but it all looks pretty decent.  I’d love to hear comments here on the methodology from competent working statisticians or econometricians.

They did a control for the idea that the change in educational outcomes was due to the change in absenteeism, and concluded it wasn’t, and a similar, but I think much more tenuous, attempt to somehow proxy for a “placebo effect”, which they interpret as suggesting the placebo effect wasn’t the thing.  I’m not so sure it matters, if it continues to work.

Nice stuff; let’s hope for expanded programs and more studies of them.  Meanwhile I think it’ll influence me to eat a little healthier myself.

Niall Ferguson is on about deficits. The Newsweek piece is full of fearmongering and frank imperialism—his biggest worry is that “economic weakness is endangering our global power”.  Well, I guess US global power, despite plenty of abuses, is more benign than most other varieties I could imagine becoming more prominent as it ebbed (except perhaps EU global power)…and could conceivably become more benign if Obama’s election ushered in an extended era of sensible, moderately liberal, multilateral policies with the world’s wellbeing among their major goals.  But the deficit fearmongering borders on the ludicrous, and while he brings in some reasonable points, the quantitative evidence is often missing and the economic reasoning slipshod or missing too.  What is perhaps most seriously lacking is a reasonable degree of balance between long-term deficit concerns and the need for deficits in the short term.

Let’s take the seriously misleading stuff first.  The US economic stimulus is described as “muted” because:

what makes a stimulus actually work is the change in borrowing by the whole public sector. Since the federal government was already running deficits, and since the states are actually raising taxes and cutting spending, the actual size of the stimulus is closer to 4 percent of GDP spread over the years 2007 to 2010—a lot less than that headline 11.2 percent deficit.

This verges on incomprehensible.  First, what is he doing with “spread over the years 2007 to 2010″?  The stimulus bill was passed in early 2009; spreading it over 2007 and 2008, years in which it was not being spent, to get an apparent lower percentage of GDP, is hogwash.  On the other hand, if we take .787 trillion as the size of the stimulus package, and spread it over 2009 and 2010, estimating 2009 GDP at 14.2 trillion and 2010 GDP at 14.697 trillion (a 3.5% growth rate, based on not much but faster than the current pace of recovery; it makes little difference to this calculation), you get a stimulus of about 2.7% of GDP.  That’s actually quite a bit less than the 4% Ferguson uses.  But that’s a big boost to aggregate demand. Not as big as it should be given the magnitude of the economic shock we’ve been hit with, but far from negligible.

I have no idea why Ferguson feels the need to start with the size of the deficit as a candidate for the size of the stimulus, and then cut it down because “we were already running deficits” and because “state and local governments are cutting taxes and raising spending”.  What matters is the change in the amount of government spending, relative to no stimulus policy.  Ferguson would appear to be overestimating the stimulus, perhaps because he’s focusing on the idea that “the deficit is the stimulus” and then “correcting” it—but a recession usually increases deficits automatically by decreasing tax receipts, without (usually) a corresponding cut in spending.  (Many states, though, now do cut spending in response to recession, often because of balanced budget laws.)  So perhaps by doing this he is including some of that recession-induced deficit in his estimate of “the stimulus”…and then reducing his overestimate by spreading it over the two prior years, as well.

The main thing that is misleading, though, is that then he estimates “the cost of this muted stimulus” by using the full deficit.  A recession is going to have a fiscal cost even without stimulus—chalking up the full 11.2% of GDP deficit as the cost of a stimulus, whether 2.7% or 4%, is either dishonest or a mistake that is really hard to excuse in an article for a national publication like Newsweek.  The fact that he uses 4% as the stimulus figure perhaps results from some mixed-up attempt to be a little less than dumb about this while still using the full deficit as the cost—perhaps this can be thought of as a way of including “automatic stabilizers” in federal spending as “stimulus”…but then arbitrarily damping it down by dividing by the extra years.  Anyway, the bottom line is that the variable spending under control of the federal government was increased by 2.7% of GDP, and roughly speaking that 2.7% is the additional cost in current dollars of the additional fiscal stimulus.  Actually, the cost is slightly less, for reasons I’ll explain shortly.  11.2% reflects the “fiscal” cost of the recession including, but not limited to, the additional stimulus spending.  A serious macroeconomic estimate of the fiscal cost of stimulus actually needs to take into account that, if you believe the economic models used to justify the stimulus itself, the stimulus increases GDP and so increases tax revenues, and so partially pays for itself.  I haven’t done this calculation—lots of people were doing it this spring, and with reasonable figures for multipliers and tax rates—say 20% for taxes and a somewhat generous 1.4 for the spending multiplier—you’d get around a 28% reduction in the total fiscal cost of the stimulus, to around 2% of GDP.  And let’s not forget that this fiscal cost is incurred in order to obtain an increase in GDP—with this multiplier, of 3.78% of GDP.  Let’s also not forget that this fiscal cost is not a loss of GDP—it is just an increase in government debt.  The GDP gain should be around a full 3.78%—perhaps less, if the stimulus is spread out longer, or the multiplier a bit lower (which it might be because some determinants of spending, like the propensity of businesses to invest, and perhaps of households to consume, may be lower than usual in the current climate of fear), but not less by the fiscal cost.

So, the 11.2% of GDP estimated federal deficit for 2009 (I should check that this is the change in net, rather than gross, public debt) is just not the cost of the “muted” stimulus, in any sense.  The “muting” of the stimulus by state budget cuts, etc., is of course argument for more short-term stimulus—for example, aid to states so they don’t have to make those budget cuts.

“We are, it seems, having the fiscal policy of a war, without the war.”  Well, hurrah for that, I say.  Better to avert a potential depression, and mitigate a serious recession, with war-footing fiscal policy than to get out of a depression the way we did in the 1940s—with the fiscal policy of an actual world war.  Of course we have the Afghan war, and had the Iraq war.  Ferguson pooh-poohs these as contributors to the fiscal situation—and here’s the second highly misleading bit of his article.

these are trivial conflicts compared with the world wars, and their contribution to the gathering fiscal storm has in fact been quite modest (little more than 1.8 percent of GDP, even if you accept the estimated cumulative cost of $3.2 trillion published by Columbia economist Joseph Stiglitz in February 2008).

Since, despite the “even if” which suggests he doubts these estimates, no other estimates are offered, let’s with Ferguson accept Stiglitz’ ones, which seem in line with what I recall hearing.  Where the heck does he come up with “little more than 1.8 percent of GDP”?  Recall that (based on three quarters of official government estimates) I estimated 209 GDP at $14.2 trillion; the total cost of the war comes to 22.5% of this year’s GDP!  (Just to clarify: this is not the yearly rate of war spending as a percentage of GDP; but Ferguson is comparing national debt to GDP, so we are looking at war debt on the same footing he’s using for total debt.) More to the point, the Nov. 25th, 2009 net federal public debt was $7.612 trillion; the estimated cost of the wars thus comes to 42%—nearly half—of the current public debt!!!  If you take at face value the CBO deficit estimates for 2019 that Ferguson cites (I have not evaluated them myself), those 3.2 trillion are still 22.4% of the projected 2019 debt of $14.3 trillion—hardly negligible (and they’d look like a higher percentage, if debt service on them were included).  I’m guessing the 1.8 percent Ferguson refers to must mean the debt service on the Stiglitz estimate of the cost of the war.  But then, since the entire rest of the public debt is less than one and a half times this war cost estimate, why doesn’t Ferguson tell us that the cost of servicing it is currently modest, too?

Now, I haven’t looked carefully, recently, into the contribution of the tax cuts of the Bush years to the deficits.  Early in the administration, they probably helped stimulate the economy out of the post-9/11 recession; but on balance they’ve added plenty to the deficit and the national debt.  By the “radical fiscal reform” we need to forestall the “fatal arithmetic of imperial decline”, does Ferguson mean things like the repeal of these tax cuts, and maybe even some modest tax hikes?  And could the current low rates private investors afford the US treasury even on long term borrowing, suggest that just maybe, these investors are pretty confident that the US will, eventually, through some combination of health care reform, tax cut repeal, tax increases, and other measures possibly including non-draconian social security adjustments, bring things into reasonable shape in the medium term, averting Ferguson’s calamitous projections?  Perhaps these low rates are a vote of confidence in a Democratic administration, as Democratic administrations have added to the national debt at a substantially lower rate than Republican ones in the past few decades. Numbers along these lines, and more on Ferguson’s article, to come.

But I can’t end without mentioning Ferguson’s comparisons to he fiscal situation of Habsburg Spain in the 16th-17th century, or prerevolutionary France—well, this sounds fun, and there may even be lessons from this far back in history.  Like, maybe it’s better to ditch the empire than trash your economy to support it?  I’m not ready to argue historical points with a historian, but it does seem like these comparisons just might be stretching it a bit, as far as the political economy of the situation and the economic and financial policy tools and knowledge available.  I do think the Habsburg Spain analogy might have fit better under Bush and Cheney.

It’s hard to beat the value and quality you get by buying the wines of a European wine-producing country in that country…especially if the country is Spain, France, or Italy.  High quality wines from smaller producers who have no need to spend time marketing their wine for export—and splitting the profits with exporters—are available, and since wine is considered a basic foodstuff, usually taxed far less than if they are exported to the US.  And most wines are the better for not having taken a boat trip across the Atlantic, or even (if you are on the West coast of North America) through the Panama Canal.

This summer I was in Spain at the biennial Quantum Information workshop in Benasque, and had the chance to try some excellent wines, and excellent values for an American even with weak dollar.  Raimat’s Viña 32 Cabernet Sauvignon 2005, DO Costers del Segre, from the inland Catalan area of Lleida (aka Lérida, in Castilian Spanish) was “excellent, with red fruits, mildly grainy tannins, fruity, slightly foxy.”  It was in the 5-6 Euro range.  This was a solid value, but a better wine was Bodegas Pirineos Merlot-Cabernet 2003 Crianza–Somontano, €8.50 at El Veedor de Viandas in Benasque.  This is from the foothills of the Pirineos near Benasque, and was a very nicely balanced, medium-bodied, easy-to-drink but not too simple wine, with a “combination of red fruits and herbs of the Pyrenees (fantasy?)”.  It lacked the “foxiness” mentioned in the Raimat, and the herbal notes leaned toward rosemary, savory, thyme, while otherwise the wine displayed typical cabernet and merlot tastes with a bit of backbone from tannin.

There are plenty of excellent cheeses to be had in Spain—just try out promising-looking ones, especially ones from local producers, in your local cheese store or supermarket cheese section and you’ll likely do fine.  One of the several that we tried with these wines was a particular standout.  Quesos de Radiguero’s Rio Vero Leche de Cabra, Moho Azul.   This is from Adahuesca in the region of Huesca, a valley or two over from Benasque but futher down, in the mid-altitude foothills of the Pyrenees.  “Moho azul” means “blue mold”, not “blue mojo”, but mold and mojo are exactly what this cheese has:  it’s covered by a finely filamented, live-looking blue-black mold, glistening with drops of moisture.  Inside it’s white to lightly straw-colored, parts smoothly creamy but parts with a slight texture like fine ricotta, balanced between runny and firm, with strong, clear, fermented or still-fermenting flavors, perhaps somewhat reminiscent of a creamy Saint Andre but more pungent, almost with hints of Roquefort.  Serious stuff, and seriously fun to scare the squeamish with.   The previous time I was in Benasque, four or so years ago, Robin Blume-Kohout told me that he considers himself a fan of strong, and blue, cheeses, but that the cheese I served at a party that summer was the first cheese he’d encountered that might possibly count as a little too strong, or too blue, for him.  Memory tells me that this cheese, which might have been Cabrales, was wrapped in leaves, something you only find in Spain anymore as it is apparently not legal for the commercial product anymore (EU regulations, no doubt).  Sergio Boixo had some other stuff to tell me about Cabrales production, but  I suspect he was pulling a prank on a credulous foreigner, so I won’t repeat it here.

Via 1Winedude, and somewhat in a similar slightly goofy (sorry, “rad”) spirit, The Second Glass Annual Wine Guide looks reasonably interesting, and I have to admit I laughed out loud at some of the fairly juvenile humor.  But some of their wine picks—Castello di Nipozzano’s Chianti Rufina (”Drink now or give to your wino friends for aging”), a couple of Proseccos (Zardetto NV, and the $12 Mionetto “Il Prosecco” with the ultra-cool beer/soda style bottle cap, on a perusal of the first few pages—Dry Creek’s Chenin Blanc (although I haven’t tried this vintage), further on) jive with mine, so I suspect this is a useful resource for wines with a good price/value ratio.  They do take ads, however, with some full-pagers from well-reviewed wines, so although I suspect they rate the wines and then drum up the ads, use your judgement.  May be aimed at the sophisticated-frat-boy market segment…Trimbach’s Pinot Blanc’s  “crisp flavors of citrus fruit and bright acidity pair perfectly with poolside hotties,”  and a Japanese wine cooler named “Saké2me” is reviewed.  And I just noticed the descriptive icons: “thanksgiving wine” … “winter warmer” ..”pizza” …what the heck is a “gift wine”…and…

Use at your own risk but this looks like it may have some useful information.

A few talks you might consider watching online:

Today (now yesterday, actually)’s colloquium.  Patrick Brady gives a nice accessible introduction to gravity waves and the current LIGO observation program, as well as what we may see when Advanced LIGO comes on line in a few years.  Some enlightening animations.  Recommended for: General Audiences. Mild equations.

Very cool ideas, very well presented, from Stephen Bartlett yesterday in the PIQuDos (Perimeter Institute Quantum Discussions series) on how the ability to do quantum computational gates—in this case, single qubit gates—in a measurement-based quantum computational model using the ground state of a lattice Hamiltonian as a resource—can in some cases be modeled by an order parameter, the expectation value of a “string” of operators on adjacent lattice sites that are related to the measurements you need to do to effect the gate.  Phases—regions of parameter space in the parametrized family of Hamiltonians—exist for which the order parameter indicates these gates work well for measurement-based QC purposes.  In particular, the qubit has to be “moved” through the lattice as a “circuit” is simulated, and in a “good” phase, the fidelity of the state of the moved qubit to the state before it’s moved, is independent of how far it has to be moved.   An excellent talk.  Recommended for: quantum computation wonks, condensed matter physicists, and a general physics audience interested in getting an idea of the cutting edge of interaction between quantum information/computation and condensed matter physics.  If you’re looking for interesting but possibly hard open problems, this talk certainly suggests some.

Here’s a wine available (as of a few weeks ago) at LCBO that I can wholeheartedly recommend:  the 2006 Nipozzano  Chianti Rufina Riserva, produced by the Marchesi de Frescobaldi.  (Don’t know what LCBO is?  Lucky you.  But they deserve some credit for stocking this.)  This is the third vintage I’ve had of this wine, and I’ve been happy with all of them, but this one’s the best.  Dark fruits, soft full flavors but definite tannic backbone to keep it together, some nice chalky minerality behind it all showing up on the longish finish with some slight hints of mintiness, and hints of bitterness.  A clear, clean, leafy version of  “Tuscan funk” lurks barely perceptible behind this, perhaps ready to contribute some heady, perfumy, but unpredictable, notes with age.  In short, fairly complex, and mixing classic Chianti characteristics with notes–including the slightest, but pleasant, hints of greenness or vegetality—that are definitely characteristic of Cabernet.  I could imagine this wine repaying 5, perhaps even 10, years of cellaring by evolving into something stunning, but this is always unpredictable, the more so as I don’t have any experience cellaring this wine.  (Comments invited from anyone who does.)

Chianti Rufina is, if I recall correctly, one of three main zones for higher-end Chiantis, the largest being Chianti Classico covering a hilly area between Florence and Siena, with Chianti Montalbano and Chianti Rufina, each much smaller than the Classico area, being the other significant DOCG’s. (I’ve also had very good Chianti from the Colli Fiorentini, i.e. the “Florentine Hills”.)

Victor Hazan’s superb, extremely well written and observed 1982 book “Italian Wine” (yes, he’s Marcella’s husband) reports Frescobaldi as “the most celebrated producer” of the area, with the Nipozzano Chianti “nearly as fine” as their single-vineyard, limited production Montesodi, but “much more accessible in price and quantity”.  At $21.55 Canadian at LCBO, that “accessible in price” still seems right 27 years on, given the quality of the wine.  I acknowledge that’s a heck of a lot to pay for a bottle if you’re not a wine geek like me, but I’m much happier paying it for this wine than the $15 to $20 CDN I’ve paid for many a mediocre bottle from LCBO (or the $10-15 US I’ve paid for some mediocre bottles in the States).

I’ll probably raise expectations too high if I quote Victor Hazan further on Chianti Rufina: “A choice Rufina can match in authority, and sometimes surpass, Chianti Classico at its finest.    In character it is closest to a Chianti from Radda [...], making forceful first impressions that precede layer after layer of unfolding flavor.” One of my best wine experiences ever (involving quantum physics, as well, so perhaps I’ll post about it at some point) involved a wine from Radda, the Monte Vertine Riserva (I don’t even recall seeing the word Chianti on the bottle), from the early 1980s.  1981 sticks in mind but at this remove—I had the wine in Turin in 1995 or so—who knows.  And this wine, though less aged, reminds me of how that Monte Vertine might have tasted in its youth.  So I think Hazan is right on in this comparison, and it says good things.  Now this is just wine, for chrissake—if you want revelation, for less money you could go out and buy the remastered deluxe edition of John Coltrane’s A Love Supreme.   But still, good stuff.  If you have the money, and the inclination, and you’ve already got a copy of A Love Supreme, give it a try.  OK, try it even if you don’t have a copy of A Love Supreme—but you really should get one of those, too.

Notes on previous editions of this wine:

2001, half bottle with dinner at Mövenpick, Zurich airport, August 2005:  “Excellent—has dark fruits and some complexity/silkiness.  Balanced.”

2002 (tasted 2004 or 2005): “Even better than the 2001, probably.  Velvety, fairly rich, notes of cocoa in the nose.  Good with George’s deep fried “little pizzas” from Campania, with red pepper, cayenne, tomato sauce.  Stands up to it.  Hints of minerality.  Superb!”  George is my son;  he likes to cook on occasion, especially Italian.  Perhaps it’s his Italian heritage from my wife, who is 100 percent Italian-American; perhaps it’s  his food-obsessed (though not more so than your average Italian) heritage from my side of the family.