Data Source

Syllabus, slides, and videos of the lectures, here you go.

China’s 89 unicorns (startups valued at $1bn or more) are worth over $350bn, by one recent estimate, approaching the combined valuation of America’s (see chart 2). And to victors go great spoils. There are 609 billionaires in China compared with 552 in America.

That is from The Economist, mostly about Chinese innovation.

From my email:

Hi, Mr. Cowen. I recently read The Complacent Class recently and enjoyed it. I’m writing because there’s an another example of American complacency that’s only come to light in recent weeks…

Specifically: the Billboard music charts..

Shape of You by Ed Sheeran last week broke the record for most weeks in top 10, with 33 weeks. The song it beat, Closer by The Chainsmokers and Halsey, set the previous record less than a year ago. http://www.billboard.com/articles/columns/chart-beat/7948959/ed-sheeran-shape-of-you-record-most-weeks-top-ten

(And yet another song in last week’s top 10, That’s What I Like by Bruno Mars, currently holds the 8th-longest record on that metric — and potentially still rising.)

Meanwhile, Despacito by Luis Fonsi, Daddy Yankee, and Justin Bieber tied the all-time record with its 16th week at #1: http://www.billboard.com/biz/articles/news/record-labels/7942315/luis-fonsi-daddy-yankee-justin-biebers-despacito-ties-for

Meanwhile, the biggest country song in the nation right now, Body Like a Back Road by Sam Hunt, is currently in its record-extending 30th week at #1 on the Hot Country Songs chart: http://www.billboard.com/files/pdfs/country_update_0905.pdf

This did not happen in decades past. Look at the Billboard charts from the ’80s — it was a new #1 song almost every week!    https://en.wikipedia.org/wiki/List_of_Billboard_Hot_100_number-one_singles_of_the_1980s

Just like how you describe in your book how people are moving less and want to stay in the same town where they were before, or how they’re switching jobs less and want to stay in the same job where they were before, people apparently just want to listen to the same songs they’ve been listening to already.

That is from Jesse Rifkin, who is a journalist in Washington, D.C. who writes about Congress for GovTrack Insider and about the film industry for Boxoffice Magazine.  Jesse sends along more:

And if you want links for statistical evidence, here are two — one about which movies have spent the most weekends in the box office top 10, the other about which songs have spent the most weeks in the Billboard top 10:

The co-authors on this paper (pdf) are Andrew Leigh and Mike Pottenger, here is the abstract:

The paper estimates long run social mobility in Australia 1870–2017 tracking the status of rare surnames. The status information includes occupations from electoral rolls 1903–1980, and records of degrees awarded by Melbourne and Sydney universities 1852–2017. Status persistence was strong throughout, with an intergenerational correlation of 0.7–0.8, and no change over time. Notwithstanding egalitarian norms, high immigration and a well-targeted social safety net, Australian long-run social mobility rates are low. Despite evidence on conventional measures that Australia has higher rates of social mobility than the UK or USA (Mendolia and Siminski, 2016), status persistence for surnames is as high as that in England or the USA. Mobility rates are also just as low if we look just at mobility within descendants of UK immigrants, so ethnic effects explain none of the immobility.

Social mobility is indeed difficult to pull off.  Hat tip goes to Ben Southwood.

They used to say this couldn’t be done:

We construct genomic predictors for heritable and extremely complex human quantitative traits (height, heel bone density, and educational attainment) using modern methods in high dimensional statistics (i.e., machine learning). Replication tests show that these predictors capture, respectively, ~40, 20, and 9 percent of total variance for the three traits. For example, predicted heights correlate ~0.65 with actual height; actual heights of most individuals in validation samples are within a few cm of the prediction. The variance captured for height is comparable to the estimated SNP heritability from GCTA (GREML) analysis, and seems to be close to its asymptotic value (i.e., as sample size goes to infinity), suggesting that we have captured most of the heritability for the SNPs used. Thus, our results resolve the common SNP portion of the “missing heritability” problem – i.e., the gap between prediction R-squared and SNP heritability. The ~20k activated SNPs in our height predictor reveal the genetic architecture of human height, at least for common SNPs. Our primary dataset is the UK Biobank cohort, comprised of almost 500k individual genotypes with multiple phenotypes. We also use other datasets and SNPs found in earlier GWAS for out-of-sample validation of our results.

While I don’t find “within a few centimeters” to be especially impressive, the question is still “what’s next?”

The authors on the paper are Louis Lello, Steven G Avery, Laurent Tellier, Ana Vazquez, Gustavo de los Campos, and Stephen D. H. Hsu.

…teenagers are increasingly delaying activities that had long been seen as rites of passage into adulthood. The study, published Tuesday in the journal Child Development, found that the percentage of adolescents in the U.S. who have a driver’s license, who have tried alcohol, who date, and who work for pay has plummeted since 1976, with the most precipitous decreases in the past decade.

The declines appeared across race, geographic, and socioeconomic lines, and in rural, urban, and suburban areas.

…Between 1976 and 1979, 86 percent of high school seniors had gone on a date; between 2010 and 2015 only 63 percent had, the study found. During the same period, the portion who had ever earned money from working plunged from 76 to 55 percent. And the portion who had tried alcohol plummeted from 93 percent between 1976 and 1979 to 67 percent between 2010 and 2016.

Teens have also reported a steady decline in sexual activity in recent decades, as the portion of high school students who have had sex fell from 54 percent in 1991 to 41 percent in 2015, according to Centers for Disease Control statistics.

Teens have also reported a steady decline in sexual activity in recent decades, as the portion of high school students who have had sex fell from 54 percent in 1991 to 41 percent in 2015, according to Centers for Disease Control statistics.

Here is the Tarah Barampour WaPo story.  Is it evolutionary psychology pushing us more into a more stable mode of behavior for safe circumstances, or perhaps teens being more aware of the need to build their resumes?  Or something else altogether different?

These developments are mostly positive, both as symptoms and as active causal agents, and yet…

Somewhere along the line there is a positive social payoff from risk-taking, including sometimes from teenagers.  How would rock and roll evolved in such a world?   Who is to help undo unjust social structures?  The graybeards?

Via Ben Schmidt, the term becomes common only in the 1970s:

I’d like to see a detailed look at actual journal practices, but my personal sense is that editorial review was the norm until fairly recently, not review by a team of outside referees.  In 1956, for instance, the American Historical Review asked for only one submission copy, and it seems the same was true as late as 1970.  I doubt they made the photocopies themselves. Schmidt seems to suggest that the practices of government funders nudged the academic professions into more formal peer review with multiple referee reports.

Further research is needed (how about we ask some really old people?), at least if peer review decides it is worthy of publication.  Frankly I suspect such work would stand a better chance under editorial review.

In the meantime, here is a tweet from the I didn’t know she was on Twitter Judy Chevalier:

I have just produced a 28-page “responses to reviewer and editor questions” for a 39-page paper.

I’d rather have another paper from Judy.

By the way, scientific papers are getting less readable.

This is from a job market paper at Stockholm University, by Sirus Dehdari:

This paper studies the effects of economic distress on support for far-right parties. Using Swedish election data, I show that shocks to unemployment risk among unskilled native-born workers account for 5 to 7 percent of the increased vote share for the Swedish far-right party Sweden Democrats. In areas with an influx of unskilled immigrants equal to a one standard deviation larger than the average influx, the effect of the unemployment risk shock to unskilled native-born workers is exacerbated by almost 140 percent. These findings are in line with theories suggesting that voters attribute their impaired economic status to immigration. Furthermore, I find no effects on voting for other anti-EU and anti-globalization parties, challenging the notion that economic distress increases anti-globalization sentiment. Using detailed survey data, I present suggestive evidence of how increased salience of political issues related to immigration channels economic distress into support for far-right parties, consistent with theories on political opportunity structure and salience of sociocultural political issues.

Here is Dehdari’s cv, all via Matt Yglesias.

There is a hot hand after all

by on September 16, 2017 at 11:21 am in Data Source, Sports | Permalink

This paper, “The Hot-Hand Fallacy: Cognitive Mistakes or Equilibrium Adjustments? Evidence from Major League Baseball,” delivers on both the theory and the empirics:

We test for a “hot hand” (i.e., short-term predictability in performance) in Major League Baseball using panel data. We find strong evidence for its existence in all 10 statistical categories we consider. The magnitudes are significant; being “hot” corresponds to between one-half and one standard deviation in the distribution of player abilities. Our results are in notable contrast to the majority of the hot-hand literature, which has generally found either no hot hand or a very weak hot hand in sports, often employing basketball shooting data. We argue that this difference is attributable to endogenous defensive responses: basketball presents sufficient opportunity for transferring defensive resources to equate shooting probabilities across players, whereas baseball does not. We then develop a method to test whether baseball teams do respond appropriately to hot opponents. Our results suggest teams respond in a manner consistent with drawing correct inference about the magnitude of the hot hand except for a tendency to overreact to very recent performance (i.e., the last five attempts).

That is from Brett Green and Jeffrey Zwiebel, via Rolf Degen.  Here are ungated versions.

We find that staggering SNAP benefits throughout the month leads to a 32 percent decrease in grocery store theft and reduces monthly cyclicity in grocery store crimes.

That is by Jillian B. Carr and Analisa Packham (pdf), via Alexander Berger.

Possibly so, though some more good years would be nice, to say the least.  To some extent this could be noise, or delayed catch-up growth.  Still, there seems to be a break in the previous trend:

In 2015, median household incomes rose by 5.2 percent. That was the fastest surge in percentage terms since the Census Bureau began keeping records in the 1960s. Women living alone saw their incomes rise by 8.7 percent. Median incomes for Hispanics rose by 6.1 percent. Immigrants’ incomes, excluding naturalized citizens, jumped by over 10 percent.

The news was especially good for the poor. The share of overall income that went to the poorest fifth increased by 3 percent, while the share that went to the affluent groups did not change. In that year, the poverty rate fell by 1.2 percentage points, the steepest decline since 1999.

…The numbers for 2016 have just been released by the Census Bureau, and the trends are pretty much the same. Median household income rose another 3.2 percent, after inflation, to its highest level ever. The poverty rate fell some more. The share of national income going to labor is now rising, while the share going to capital is falling.

That is from the new David Brooks column.

That topic has been knocking around for some time, with varying opinions.  I’ve now seen the clearest and most thorough treatment to date, namely from Gerald Auten and David Splinter.  It hasn’t received that much attention, perhaps because the results don’t have such a strong built-in constituency, but here goes:

Previous studies using U.S. tax return data conclude that the top one percent income share increased substantially since 1960. This study re-estimates the long-term trend in inequality after accounting for changes in the tax base, income sources missing from individual tax returns and changes in marriage rates. This more consistent estimate suggests that top one percent income shares increased by only about a quarter as much as unadjusted shares. Further, accounting for government transfers suggests that top one percent shares increased a tenth as much. These results show that unadjusted tax return based measures present a distorted view of inequality trends, as incomes reported on tax returns are sensitive to changes in tax laws and ignore income sources outside the individual tax system.

You’ll find the paper at the first link here.

In a new NBER working paper David Card and Abigail Payne have a stunning new explanation of the gender gap in STEM at universities. The conventional wisdom is that the gender gap is about women and the forces–discrimination, sexism, parenting, aptitudes, choices; take your pick–that make women less likely to study in STEM fields. Card and Payne are saying that the great bulk of the gap is actually about men and their problems. At least that is my interpretation of their results, the authors, to my mind, don’t clearly state just how much their results run against the conventional wisdom. (Have I misunderstood their paper? We shall see.)

The authors are using a large data set on Canadian high school students that includes data on grade 12 (level 4) high school classes and grades and initial university program. Using this data, the authors find that females are STEM ready:

…At the end of high school, females have nearly the same overall rate of STEM readiness as males, and
slightly higher average grades in the prerequisite math and science courses.  The mix of STEM related courses taken by men and women is different, however, with a higher concentration of women in biology and chemistry and a lower concentration in physics and calculus.

Since females are STEM-ready when leaving high school you are probably thinking that the gender gap must be a result either of different entry choices conditional on STEM-readiness or different attrition rates. No. Card and Payne say that entry rates and attrition rates are similar for males and females. So what explains why males are more likely to take a STEM degree than females?

The main driver of the gender gap is the fact that many more females (44%) than males (32%) enter university.  Simply assuming that non‐STEM ready females had the same university entry rate as non‐STEM ready males would
narrow the gender gap in the fraction of university entrants who are STEM ready from 14
percentage points to less than 2 percentage points.

Moreover:

On average, females have about the same average grades in UP (“University Preparation”, AT) math and sciences courses as males, but higher grades in English/French and other qualifying courses that count toward the top 6 scores that determine their university rankings. This comparative advantage explains a substantial share of the gender difference in the probability of pursing a STEM major, conditional on being STEM ready at the end of high school.

Put (too) simply the only men who are good enough to get into university are men who are good at STEM. Women are good enough to get into non-STEM and STEM fields. Thus, among university students, women dominate in the non-STEM fields and men survive in the STEM fields. (The former is mathematically certain while the latter is true only given current absolute numbers of male students. If fewer men went to college, women would dominate both fields). I don’t know whether this story will hold up but one attractive feature, as a theory, is that it is consistent with the worrying exit from the labor market of men at the bottom.

If we accept these results, the gender gap industry is focused on the wrong thing. The real gender gap is that men are having trouble competing everywhere except in STEM.

Hat tip: Scott Cunningham.

I haven’t had a chance to look at this one, but here is the headline summary from Brookings:

The new paper, published in the Fall 2017 edition of the Brookings Papers on Economic Activity, makes a strong case for looking at the opioid epidemic as one driver of declining labor force participation rates.

In fact, Krueger suggests that the increase in opioid prescriptions from 1999 to 2015 could account for about 20 percent of the observed decline in men’s labor force participation during that same period, and 25 percent of the observed decline in women’s labor force participation.

Here is the Brookings link.