Data Talks

I was writing a post on standardized tests and I thought I'd share one of the graphs

Here's the question: how many SAT points can parents buy their kids?

It's a hot topic: people claim wealth can *buy* a kid's high scores. Reality disagrees

300 hours of coaching gets you <1 SD pic.twitter.com/iWzO1G1RCd

— Crémieux (@cremieuxrecueil) August 17, 2023

 

We are living with the consequences of this occlusion.

From My Mother Survived the Nazis.  My Father Survived the Soviets. by Daniel Finkelstein.  The subheading is Yet the U.S.S.R. was never held accountable for its war crimes.

It has always puzzled me how reluctant American academia has been to draw parallels between National Socialism (Nazis) and the Union of Soviet Socialist Republics (Soviet Union).  Both were nationalistic, militaristic, authoritarian antitheses of Clasical Liberalism.  Sure, the Soviet Union was an alliance of convenience in a divide-and-conquer strategy, but they were each in their fashion just as much an enemy of virtually every tenet of Age of Enlightenment Classical Liberalism.

Even still, in academia, there is a touching naive faith in the vanguard of socialism and the power of central planning.  We call it ESG or DEI or Critical Race Theory or Social Justice or postmodernism now, but it is still all the same old anti-Classical Liberalism heart throb for punitive authoritarianism.  

Finkelstein's is one family's special fate to have been simultaneously the victims of both Naziism and Soviet Communism.  

“Should i mention that I saw Anne Frank in Belsen? Do you think they’d be interested in that?” I was in my late teens when my mother was first asked to give a talk about her experiences as a German refugee and Dutch Jew in the Second World War. Until the late 1970s, people rarely asked her about it, and she didn’t want to be a bore.

Then things began to change. Within a few years of her first speech, she was giving lectures in schools quite regularly. She was invited to Downing Street and talked with the prime minister about knowing the Franks, and about her father’s work fighting fascism and his encounter with Hermann Göring. The BBC made a documentary in which my mother met the daughter of a prominent Nazi. She was forever telling her story.

No one, however, ever asked Dad to tell his. The interest in what happened to him never came. It still hasn’t come.

Yet my father was the victim of one of the war’s greatest crimes: Stalin’s attempt to eradicate the Polish nation by murdering its elite and scattering its leadership. It was a crime that saw hundreds of thousands of people expelled from their homes and deported to become slave laborers, and saw hundreds of thousands more imprisoned in terrible conditions. It’s a story little told, often denied, and, even now, to most people, entirely unknown. My father’s story is one that history has half hidden.

Decades later, we are living with the consequences of this occlusion.

It is an astonishing and tragic tale well worth broader circulation.  Read the whole thing.

From Social norms: the downside by David Hugh-Jones.  The subheading is "Surely under well-defined circumstances, ye CAN push yer grannie off a bus?"  An entertaining musing on social norms.  

The Yimby (“Yes In My Back Yard”) movement in the US has sprung up to fight for less restrictive zoning and more housing, often coming up against Nimby (“Not In My Back Yard”) opponents. Broadly, this is a contest of ideals against self-interest, and broadly I think the Yimbies are right. But the self-interest of Nimbies also carries some moral weight — it’s genuinely a loss to people when transport systems get more crowded, it’s even genuinely a loss when a given neighbourhood gets less “exclusive” than it was! I don’t expect Yimbies to be good at calculating the perhaps rare cases when that loss outweighs the gain.

A reasonable model of politics is that left-wing parties are norm-driven and right-wing parties are driven by self-interest. I speak as a conservative! It may really be true that, as Tony Blair once said, people on the left are “pretty straight guys”. The downside of being ideal-driven is that your activists’ ideals may not involve much cost-benefit analysis. Canny right-wing politicians can slice off chunks of their opponents’ support by embracing idealistic causes that have little or no cost, and perhaps little or no benefit. David Cameron was great at this. Gay marriage, foreign aid and global warming were all ways to rebrand the Tory party. Some of these are unimportant in the grand scheme of things. Others are important, but carried serious costs only in the post-David-Cameron future, which arrived sooner than he expected. But the left wing is also quite capable of doing this to itself. The Scottish National Party has just spent scarce parliamentary time making it easier for teenagers to choose their gender. This is a non-issue for almost every person in Scotland, but presumably it makes SNP activists feel good.

In extreme cases, cost-free norms can become really absurd. On a flight to Bombay in the 1990s, the inflight magazine had an article: Is India Racist? The investigation proceeded by sending a black guy into various high-class shops to see how he was treated. 1990s India was a country of almost a billion people. The number of Africans or African Americans, even including ethnic groups like the Siddi who have been in India for centuries, is essentially a rounding error. By contrast it contained hundreds of millions of low-caste people, facing discrimination, entrenched poverty, debt bondage, you name it; and about a hundred million Muslims, who then as now were quite often — never mind discriminated against! — persecuted and murdered. But in the very rich US, people care about anti-black racism, and it’s important to be like the very rich US. So: how did Bombay shop assistants treat a black guy? With startled amazement, I suppose.

 I consider social norms as a coding mechanism which transmits knowledge and behaviors inter-generationally, covering the gap between evolved nature and changed context and circumstances.  The gap is always there and the nature of the gap also drives a path dependency in the evolution of those social norms.

Hugh-Jones is correct that for all the benefits of social norms in individuals and collectives to adapt to changed circumstances as well as social norms which facilitate non-genetic fit for purpose knowledge and behaviors, they are not without cost.  They are usually net-beneficial but we should not simply assume into existence a net benefit which might not be there.  Context and circumstance are changing far too quickly for that assumption to rest easy. 

The last paragraph above calls to mind a piece I read in the Daily Telegraph a few days ago.  From It’s time to admit it: Black Lives Matter hysteria made fools of us all by Michael Deacon.  The subheading is As a British protest organiser admits fraud, it’s hard to help wondering: what exactly came over us in summer 2020? 

Remember the Black Lives Matter protest in Bristol? The one where the mob tipped the statue into the harbour? In court this week, one of its organisers admitted fraud, after £30,000 in donations from Black Lives Matter supporters went missing.

Those donors will be shaking their heads. But I suspect they won’t be the only ones. Because, three years on, it feels like an appropriate time to reflect on what happened during that mad summer of 2020 – and to ask: what exactly came over us?

We all recall what happened. For almost a month, mass protests raged all over Britain. Outside Downing Street, a policeman took the knee while on duty. The Labour leader, and his deputy, posed for a photograph in which they solemnly took the knee inside Parliament.

Then the England football team started taking the knee, too. In fact, they continued to perform this American gesture long after American sportsmen had stopped. Hence the peculiar spectacle before the England v USA match at last year’s World Cup. The England team took the knee – but the US team didn’t.

All of the above happened in response to a single event: the murder of an African-American man by a white American policeman in the American state of Minnesota, 4,000 miles away. The murder was of course appalling. But how come such huge protests against US cops erupted in Britain, too?

Deacon thinks the madness was a response to the tensions from Covid lockdown.  This ignores that British academia have for some decades been susceptible to parroting the more extreme social justice nonsense of their American brethren even when the fit is poor.  

Human capital (such as knowledge, skills, and values) is transmitted within families, and social capital embodied in kinship networks also survives revolutions

From Persistence Despite Revolutions by Alberto F. Alesina, et al.  From the Abstract.

Can efforts to eradicate inequality in wealth and education eliminate intergenerational persistence of socioeconomic status? The Chinese Communist Revolution and Cultural Revolution aimed to do exactly that. Using newly digitized archival records and contemporary census and household survey data, we show that the revolutions were effective in homogenizing the population economically in the short run. However, the pattern of inequality that characterized the pre-revolution generation re-emerges today. Almost half a century after the revolutions, individuals whose grandparents belonged to the pre-revolution elite earn 16 percent more income and have completed more than 11 percent additional years of schooling than those from non-elite households. We find evidence that human capital (such as knowledge, skills, and values) has been transmitted within the families, and the social capital embodied in kinship networks has survived the revolutions. These channels allow the pre-revolution elite to rebound after the revolutions, and their socioeconomic status persists despite one of the most aggressive attempts to eliminate differences in the population.

Fits with Gregory Clark's work.  

Upper class people are not getting married to lower class people.

From Response to Douthat on Two-Parent Families by Matt Bruenig.  He makes a good point which I had not considered and which is germane in terms of policy considerations.  He is discussing marriage, class and income/well-being.  Read the whole thing but the gist is this.

Regarding Douthat’s marriage point here, he makes a mistake that nearly everyone in this discourse makes and it’s kind of driving me crazy over time. The mistake is saying that, contrary to lower class people, upper class people are getting married. The reason this is a mistake is that upper class people are not getting married to lower class people.

If we follow Douthat’s recommendation to look at the behavior of upper class people to determine what kinds of behavior are privilege-enhancing, then we are forced to conclude that marrying an upper class spouse is privilege-enhancing while marrying a lower class spouse is not and should be avoided. That’s how upper class people actually behave. Right?

Another common and pithy way of expressing Douthat’s point here is that upper class people should “preach what they practice,” with proponents of that phrase seemingly thinking that it means that upper class people should tell lower class people to marry. But what upper class people practice is not “marriage.” It’s “marriage to upper class people.” Right?

Indeed, if you want an example of how a certain marriage proponent approached marriage in her own life, she conveniently had the New York Times announce the details of her marriage in 2001.

Melissa Jean Schettini, a daughter of Mr. and Mrs. Louis A. Schettini of Montville, N.J., was married yesterday to Daniel Patrick Kearney Jr.

The bride and bridegroom, both 27, met at Princeton University, where they graduated, she summa cum laude.

Mrs. Kearney is a candidate for a Ph.D. in economics at the Massachusetts Institute of Technology,

Her father owns Per Forms, a Montville company, which manufactures and markets business forms.

Next month, Mr. Kearney is to become a law student at Yale. He received a master’s degree in classical studies from Boston College.

The bridegroom’s father retired as the chief investment officer of Aetna, the insurance company in Hartford. He was also the president of its investments and financial services division.

Is the upshot from this kind of behavior really “people should get married”? Or is it “when the upper class boyfriend gets the letter from Yale law school, go ahead and lock that down”?

Is it offensive to write this? I gather from my surveying of the society that is. But why is it offensive? Is it because second-guessing people’s relationships is offensive? And if so, what should we make of people who write whole books doing that?

I take his point.  I remain convinced there are, on average, pro-social benefits to all when people marry.  

Are there lower probabilities or occurrence and greater probabilities of failure when it is a cross-marriage?  Cross-religion, cross-class, cross-race, cross-culture?  Sure.  We know these marriages can and do work.  We know they involve greater complexity.  They may possibly entail greater failure rates (I would default to that assumption but have not seen data on it.)  

I suspect that Bruenig is overstating just how hermetically sealed things are in terms of class.  But his general point that people assortatively mate within culture, class, education attainment, ethnicity, religion, etc. is usefully true.  And to the extent that it is strongly true, it raises legitimate questions about how transferable any policy might be across such groupings.

Reconciling claims with reality

I do a lot of consulting in the energy sector and consequently am constantly startled by the claims I frequently encounter, not just among ESG advocates, but from government agencies and the like, that renewable energy (wind, solar, tidal, biomass) is cost-competitive with traditional sources (nuclear, coal, gas, and hydro).  

This just isn't true.  There is, of course, a wrinkle.

There is the straight forward comparison of cost of production.  It is quite hard to determine what the exact number is because of so many regulations and hidden subsidies of one sort or another.  All energy sources receive subsidies and all pay taxes and absorb government costs.  You have to disentangle them and arrive at definitions everyone can agree on.

An example from transportation.  In comparing the costs of various forms of transportations, should one consider construction and maintenance of roads as a form of subsidy to the auto industry.  In some regards that is a nonsense position, though frequently taken by anti-car advocates.  On the other hand it is a fair question.  

In that example it really comes down to the will and choice of the people.  Anti-car advocates wish to use the regulatory and coercive power of government to limit people's choices and welfare.  When given the choice, people always choose the freedom, efficiency, and adaptability of the car over any other alternative.  In which case, building and maintaining roads is not a subsidy butt he provision of a collective good.  But you do need to consider the definitions and think it through.

With energy there is a similar issue.  The objective is the provision of clean, safe, cheap and reliable energy when and where it is needed in the quantities required.  There are a further two aspects of that provision, base load and peak.  The cost profiles differ between base and peak loads.

Base load is the minimum necessary MW necessary to meet the lowest demand in a period of time.  Base load plants are usually highly efficient both in terms of cost and energy conversion.  Historically the traditional sources of energy serve as base load units by their very nature.  The one drawback is that other than natural gas, base load generators take a while to prepare and be brought online.  And even gas is not a mere matter of turning a dial.  Traditional energy generation (other than hydro) can be located wherever it is needed.  

By their very nature, renewables (wind, solar, biomass, tidal) can not be located where they are most needed, nor can they be relied upon to serve as base load.  So do you compare the costs of solar generated power in peak terms against coal at peak, coal at base, or coal at average?

Because of these definitional issues as well as regulatory complexity and obscurity all further complicated by cost accounting practices which would impress Hollywood studio executives, the claims that renewable energy costs have seemed grossly suspect but not easily disprovable.

I just came across one source that indirectly helps clarify the claims.  From Federal Financial Interventions and Subsidies in Energy in Fiscal Years 2016–2022 by the US Energy Information Administration.  

They ignore the distinction between base and peak, going with average.  They focus only on federal subsidies and not State or Local subsidies which can be huge.  There are definitional arguments to be had about what counts as a subsidy but they are pretty clear in each case what they are defining as in and out.  They provide the data but not really much interpretation.  We want to know the relative cost of each energy source.  They provide the data to determine that but don't do the analysis to give us that information.

Fortunately, we have Biden admin quietly released study showing green energy receives far more subsidies than fossil fuels by Thomas Catenacci.  The subheading is Solar should be competing for sales in the marketplace, not for subsidies in Washington,' top Republican senator tells Fox News Digital.  The reporting is pugnacious but it does give the answers.

The report calculates that there were $183.3 billion in Federal subsidies between 2016 and 2022 which seems quite low to me.  We are running an average of $26 billion in subsidies for the different categories of energy generation.  

Over the seven years, energy production is up 21% across all sources.  Where do we stand today in terms of energy generation?  Where do we get the energy for the electricity we generate?  For 2022, the answers are, in order of importance.  Bold are our traditional sources and italic are the supposed waves of the future.  

Natural Gas - 43% of our energy generation (average of 41% over the seven years)

Oil - 24% of our energy generation (average of 24% over the seven years)

Coal - 12% of our energy generation (average of 15% over the seven years)

Nuclear - 8% of our energy generation (average of 9% over the seven years)

Biomass - 5% of our energy generation (average of 5% over the seven years)

Wind - 4% of our energy generation (average of 3% over the seven years)

Hydro - 2% of our energy generation (average of 3% over the seven years)

Solar - 2% of our energy generation (average of 1% over the seven years)

Geothermal - 0% of our energy generation (average of 0% over the seven years)

A couple of striking things in contrast to the general impression from headlines.  

Our source structure is very stable over the seven years.  A couple of percent up or down, here or there, but not much movement.  

For all the hoopla and subsidies, wind and solar have only gone from a cumulative 3% to 6%.  Yes, a doubling, but a doubling of nearly nothing.  There really isn't much there, there.  

That's where we get our energy from.  Now, where do the subsidies go?  If, as is often claimed, Solar and Wind are now competitive or cheaper than other sources of energy, they ought in theory not be receiving any subsidies.  In this analysis, natural gas a liquid petroleum products are combined.  For 2022.

Natural Gas & Oil - 67% of our energy generation; 15% of subsidies

Coal - 12% of our energy generation; 6% of subsidies

Nuclear - 8% of our energy generation; 3% of subsidies

Biomass - 5% of our energy generation: 2% of subsidies

Wind - 4% of our energy generation; 23% of subsidies

Hydro - 2% of our energy generation; 0% of subsidies

Solar - 2% of our energy generation; 49% of subsidies

Geothermal - 0% of our energy generation; 2% of subsidies

Our main sources of traditional energy (gas, oil, coal and nuclear), supplying 87% of all energy, receive only 24% of the subsidies on offer.  

Wind and Solar, providing 6% of our energy, still receive 72% of all subsidies.

The evidence suggests that Wind and Solar are still not viable.  And, given that they are receiving large subsidies while covering only a small percentage of energy needs, it seems unlikely that they will ever be viable under existing technology and engineering constraints.  

What we really want is an EIA report that provides the real cost, net of subsidies, for each of our energy sources.  I have not yet come across such a report.  There are older studies but we need something up to date.  

But this one from August certainly lends credence to my supposition that what is reported in the mainstream media versus what I glean from inside the industry are mismatched and that it is the mainstream media version which is not aligned with reality.

You could argue that solar and wind are cheaper than traditional energy but that the subsidies are still needed to accelerate the transition.  But we have been in heavy duty transition for fifteen years or more (who can forget the half billion dollar loss from Solyndra in 2010?)  This should be the period of picking low hanging transition fruit.  If in fifteen years and all the easy opportunities, we have a=only gone from 3% of energy to 6%, it does not bode well for the cost structure in the future.

History

A rare look at a subterranean Roman aqueduct channel, this one built by the Roman army to bring water to Viminacium in modern Serbia. Despite the popular image of grand aqueduct bridges, most Roman aqueducts ran almost entirely underground. pic.twitter.com/djU1fHYXUD

— Gareth Harney (@OptimoPrincipi) August 17, 2023

 

An Insight

Western Civilization, Chinese Style https://t.co/EiNAvSSeM4

— Charles Bayless (@CharlesBayless) August 17, 2023

 

The more confounding variables that you attempt to control for, the more you expand researcher degrees of freedom

From Fine, I'll run a regression analysis. But it won't make you happy. by Nate Silver.  The subheading is State partisanship and COVID vaccination rates are strongly predictive of COVID death rates even once you account for age.

One of my rules when I get in public debates as a statistician is: The simpler, the better. More or less, this is a version of Occam’s Razor. The more complications you introduce into an analysis, the more confounding variables that you attempt to control for, the more you expand researcher degrees of freedom — in other words, decision points by the analyst about how to run the numbers.

I don’t think it’s quite right to say these decisions are arbitrary. Ideally they’ll reflect a statistician’s judgment, experience and familiarity with the subject matter. Sometimes it’s absolutely necessary to control for confounders: otherwise you might wind up with ridiculous implications like that consuming ice cream causes drowning (summer weather is the obvious confounding variable). However, there are trade-offs when adding complications to your analysis. Although it’s possible to err in either direction, there’s a general tendency to overfit models.

My keep-it-simple attitude is also a stress response from years of experience arguing on the Internet. Any time you can make your point using simple counting statistics or other very straightforward methods, I consider that a win. People usually aren’t really interested in the intricacies beyond a certain point. Most of the time, what Scott Alexander calls “isolated demands for rigor” — there’s always some factor you haven’t accounted for — are just stepping stones on the road to confirmation bias.

So my aim is generally to focus on stylized facts that are true and robust. And to keep repeating them. I like simple (or simple-seeming) claims that — and I can’t emphasize this last part enough — I expect will hold up to scrutiny.

Related to Von Neumann's comment on curve fitting:

With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.

Another individual becoming increasingly skeptical of the data integrity of university research - Nobel prize winner Daniel Kahneman.

In an interview, Dr. Kahneman, the Nobel Prize winner, suggested that while the efforts of scholars like the Data Colada bloggers had helped restore credibility to behavioral science, the field may be hard-pressed to recover entirely.

“When I see a surprising finding, my default is not to believe it,” he said of published papers. “Twelve years ago, my default was to believe anything that was surprising.”

I see wonderful things

Adansonia grandidieri is the biggest and most famous of Madagascar's six species of baobabs. They can reach 25 to 30 m (82 to 98 ft) in height

[📹 cityscape_discovery]pic.twitter.com/P7qaUiBK9j

— Massimo (@Rainmaker1973) August 17, 2023