Month: January 2020

My Conversation with the excellent Reid Hoffman

This one is better than the other available conversations with Reid, here is the transcript and audio.  Here is part of the CWTeam summary:

Reid joined Tyler to talk about all these leverage points and more, including the Silicon Valley cultural meme he most disagrees with, how Wittgenstein influenced the design of LinkedIn, mystical atheism, what it was like being on Firing Line, why he’s never said anything outrageous, how he and Peter Thiel interpret The Tempest differently, the most misunderstood thing about friendship, how to improve talent certification, what’s needed from science fiction, and his three new ideas for board games.

Excerpt:

COWEN: If we think of Peter Thiel and Elon Musk, they could arguably, by the standards of many people, be called weird. I’ve reviewed all the books you’ve written and a lot of your public talks. I can’t recall you saying a single thing that’s outrageous in any way whatsoever. Why aren’t you weirder?

HOFFMAN: [laughs] Maybe I mask it better. That’s my Straussian element, that I hide my weirdness. I would say that a little bit of it comes down to a theory about what is the right way of evolving discourse.

I think I probably do have a variety of views that people would think is weird. I, for example, think of myself as a mystical atheist, which is neither the full atheist category nor any religious category, but some blend in the middle. Or the fact that I actually think that the notion of capitalism is one of the world’s leading interesting technologies, but it’s not a particularly good philosophy, and you’d think that’s odd for an entrepreneur or an investor, and so forth.

So I have areas where I would say groups of people would think I’m weird. I may not highlight it because I tend to always speak in a way to, how do I think I help us make the most progress? And I would only say the weird things if I thought that was the thing that would result from that.

COWEN: So there are weird things that are in your mind?

HOFFMAN: Yes, yeah.

And:

COWEN: How did your interest in the late Wittgenstein influence the construction and design of LinkedIn? I’m sure they ask you this all the time in interviews.

HOFFMAN: [laughs] All the time. The question I’ve always been expecting. I would say that the notion of thinking about — a central part of later Wittgenstein is to think that we play language games, that the way that we form identity and community, both of ourselves and as individuals, is the way that we discourse and the way that we see each other and the way that we elaborate language.

That pattern of which ways we communicate with each other, what’s the channel we do, and what’s the environment that we’re in comes from insights from — including later Wittgenstein, who I think was one of the best modern philosophers in thinking about how language is core to the people that we are and that we become.

COWEN: What else from philosophy influenced the construction and design of LinkedIn?

Recommended.  For help in arranging this Conversation I am very much indebted to Ben Casnocha.

The United States is Starved for Talent

The US offers a limited number of H1-B visas annually, these are temporary 3-6 year visas that allow firms to hire high-skill workers. In many years, the demand exceeds the supply which is capped at 85,000 and in these years USCIS randomly selects which visas to approve. The random selection is key to a new NBER paper by Dimmock, Huang and Weisbenner. What’s the effect on a firm of getting lucky and wining the lottery?

We find that a firm’s win rate in the H-1B visa lottery is strongly related to the firm’s outcomes over the following three years. Relative to ex ante similar firms that also applied for H-1B visas, firms with higher win rates in the lottery are more likely to receive additional external funding and have an IPO or be acquired. Firms with higher win rates also become more likely to secure funding from high-reputation VCs, and receive more patents and more patent citations. Overall, the results show that access to skilled foreign workers has a strong positive effect on firm-level measures of success.

Overall, getting (approximately) one extra high-skilled worker causes a 23% increase in the probability of a successful IPO within five years (a 1.5 percentage point increase in the baseline probability of 6.6%). That’s a huge effect. Remember, these startups have access to a labor pool of 160 million workers. For most firms, the next best worker can’t be appreciably different than the first-best worker. But for the 2000 or so tech-startups the authors examine, the difference between the world’s best and the US best is huge. Put differently on some margins the US is starved for talent.

Of course, if we play our cards right the world’s best can be the US best.

Sam Altman on start-up talent

I look for founders who are scrappy and formidable at the same time (a rarer combination than it sounds); mission-oriented, obsessed with their companies, relentless, and determined; extremely smart (necessary but certainly not sufficient); decisive, fast-moving, and willful; courageous, high-conviction, and willing to be misunderstood; strong communicators and infectious evangelists; and capable of becoming tough and ambitious.

Some of these characteristics seem to be easier to change than others; for example, I have noticed that people can become much tougher and more ambitious rapidly, but people tend to be either slow movers or fast movers and that seems harder to change. Being a fast mover is a big thing; a somewhat trivial example is that I have almost never made money investing in founders who do not respond quickly to important emails.

Also, it sounds obvious, but the successful founders I’ve funded believe they are eventually certain to be successful.

Here is the full blog post — agree or disagree?

The best available fix for real wage stagnation

That is the topic of my latest Bloomberg column, here is one excerpt:

In other words, the frontier areas for overcoming wage stagnation are several-fold. First is a greater freedom to build, so that housing supply can rise and prices can fall. That also would enable more upward mobility by easing moves to America’s more productive (but also more expensive) regions. Second are steps to lower the cost of medical care through greater competition and price transparency. Third, American higher education is hardly at its optimum point of efficiency, innovation and affordability.

If those sectors displayed some of the dynamism and innovativeness of that marks America’s tech sector, the combination of declining prices and rising quality could give living standards a boost. And since rent, health care and tuition tend to be higher shares of the incomes of poorer people, those changes would help poorer people the most.

Think of it as a rooftops piece, combined with a discussion of why wages actually have seen slow growth as of late.

Tuesday assorted links

1. The most checked out books of all time from the New York Public Library.  #1 is The Snowy Day, by Ezra Jack Keats.

2. Why gaming is outperforming TV, recommended.

3. Flying has become much cheaper over the last five years (sometimes trotted out as a case of “monopoly,” because the number of major carriers went down).

4. Deaflix: Deaf Entertainment Network.

5. Bagel Union > Mafia.

Big Data+Small Bias << Small Data+Zero Bias

Among experts it’s well understood that “big data” doesn’t solve problems of bias. But how much should one trust an estimate from a big but possibly biased data set compared to a much smaller random sample? In Statistical paradises and paradoxes in big data, Xiao-Li Meng provides some answers which are shocking, even to experts.

Meng gives the following example. Suppose you want to estimate who will win the 2016 US Presidential election. You ask 2.3 million potential voters whether they are likely to vote for Trump or not. The sample is in all ways demographically representative of the US voting population but potential Trump voters are a tiny bit less likely to answer the question, just .001 less likely to answer (note they don’t lie, they just don’t answer).

You also have a random sample of voters where here random doesn’t simply mean chosen at random (the 2.3 million are also chosen at random) but random in the sense that Trump voters are as likely to answer as are other voters. Your random sample is of size n.

How big does n have to be for you to prefer (in the sense of having a smaller mean squared error) the random sample to the 2.3 million “big data” sample? Stop. Take a guess….

The answer is…here. Which is to say that your 2.3 million “big data” sample is no better than a random sample of that number minus 1!

On the one hand, this illustrates the tremendous value of a random sample but it also shows how difficult it is in the social sciences to produce a truly random sample.

Meng goes on to show that the mathematics of random sampling fool us because it seems to deliver so much from so little. The logic of random sampling implies that you only need a small sample to learn a lot about a big population and if the population is much bigger you only need a slightly larger sample. For example, you only need a slightly larger random sample to learn about the Chinese population than about the US population. When the sample is biased, however, then not only do you need a much larger sample you need it to large relative to the total population. A sample of 2.3 million sounds big but it isn’t big relative to the US population which is what matters in the presence of bias.

A more positive way of thinking about this, at least for economists, is that what is truly valuable about big data is that there are many more opportunities to find random “natural experiments” within the data. If we have a sample of 2.3 million, for example, we can throw out huge amounts of data using an instrumental variable and still have a much better estimate than from a simple OLS regression.

Gangs really matter

We study the effects that two of the largest gangs in Latin America, MS-13 and 18th Street, have on economic development in El Salvador. We exploit the fact that the emergence of gangs in El Salvador was in part the consequence of an exogenous shift in US immigration policy that led to the deportation of gang leaders from the United States to El Salvador. Using the exogenous variation in the timing of the deportations and the boundaries of the territories controlled by the gangs, we perform a spatial regression discontinuity design and a difference-in-differences analysis to estimate the causal effect that living under the rule of gangs has on development outcomes. Our results show that individuals living under gang control have significantly worse education, wealth, and less income than individuals living only 50 meters away in areas not controlled by gangs. None of these discontinuities existed before the arrival of gangs from the US. The results are not determined by exposure to violence, lower provision of public goods, or selective migration away from gang locations. We argue that our findings are mostly driven by gangs restricting residents’ mobility and labor choices. We find that individuals living under the rule of gangs have less freedom of movement and end up working in smaller firms. The results are relevant for many developing countries where non-state actors control parts of the country.

That is from a new paper by Nikita Melnikov, Carlos Schmidt-Padilla, and Maria Micaela Sviatschi.  Via the excellent Samir Varma.

Urban growth and its aggregate implications

That is the title of a new paper by Gilles Duranton and Diego Puga.  This piece goes considerably beyond previous research by having a more explicit model of both urban-rural interactions, and also possible congestion costs arising from more YIMBY.  Here are a few results of the paper:

1. If you restricted New York City and Los Angeles to the size of Chicago, 18.9 million people would be displaced and per capita rural income would fall by 3.6%, due to diminishing returns to labor in less heavily populated areas.

2. The average reduction in real income per person, from this thought experiment, would be 3.4%.  You will note that NIMBY policies are in fact running a version of this policy, albeit at different margins and with a different default status quo point.

3. If you were to force America’s 11 largest cities to be no larger than Miami, real income per American would fall by 7.9%.

4. If planning regulations were lifted entirely, NYC would reach about 40 million people, Philadelphia 38 million (that’s a lot of objectionable sports fans!), and Boston just shy of 30 million (ditto).

5. Output per person, under that scenario, would rise in NYC by 5.7% and by 13.3% in Boston.  That said, under this same scenario incumbent New Yorkers would see net real consumption losses of 13%, whereas for Boston the incumbent losses are only about 1.1%.

6. The big winners are the new entrants.  On average, real income would rise by 25.7%.

7. Alternatively, in their model, rather than laissez-faire, if America’s three most productive cities relaxed their planning regulations to the same level as the median U.S. city, real per capita income would rise by about 8.2%.

8. In all of these cases the authors calculate the change in rural per capita income, based on resulting population reallocations.

Recommended, I am very glad to see more serious work in this area.

The U.S. cultural elite

Overall I do not regard this as good news:

We examine the educational backgrounds of more than 2,900 members of the U.S. cultural elite and compare these backgrounds to a sample of nearly 4,000 business and political leaders. We find that the leading U.S. educational institutions are substantially more important for preparing future members of the cultural elite than they are for preparing future members of the business or political elite. In addition, members of the cultural elite who are recognized for outstanding achievements by peers and experts are much more likely to have obtained degrees from the leading educational institutions than are those who achieve acclaim from popular audiences.

That is from a new paper by Steven Brint, et.al., via the excellent Kevin Lewis.

Monday assorted links

The culture that is Buckingham: what is the underlying structural model?

The Duke of Cambridge has spoken of his “sadness” at the broken bond with his brother and voiced sorrow that the royal family is no longer a “team”.

As the Queen called emergency peace talks tomorrow at Sandringham to end the Windsors’ civil war, The Sunday Times can reveal that Prince William has said he feels sorrow that he and Prince Harry are now “separate entities” and expressed hope that they might pull together again in future.

“I’ve put my arm around my brother all our lives and I can’t do that any more; we’re separate entities,” he told a friend.

…Tom Bradby, who did the recent ITV interview in which Harry and Meghan confessed their sense of isolation, warned failure to keep the pair on side could lead the Duke and Duchess of Sussex to do a “no-holds-barred” interview that could damage the monarchy further.

…Harry and Meghan may have their security downgraded, with protection squad officers armed only with Tasers rather than guns.

Here is the full piece from the London TimesSome reports say Meghan now has a deal with Disney, maybe she will do voice-over for a princess…

The culture that was Parfit

“‘Like my cat, I often simply do what I want to do.’ This was the opening sentence of Derek Parfit’s philosophical masterpiece, Reasons and Persons… However, there was a problem. Derek did not, in fact, own a cat. Nor did he wish to become a cat owner, as he would rather spend his time taking photographs and doing philosophy. On the other hand, the sentence would clearly be better if it was true. To resolve this problem Derek drew up a legal agreement with his sister, who did own a cat, to the effect that he would take legal possession of the cat while she would continue living with it.”

And this one:

Derek Parfit was famously a fast and creative thinker. He used to advise students and colleagues to set up autocomplete shortcuts on MS Word for their most commonly used phrases to boost their productivity, unaware that very few other philosophers felt that their productivity was being restricted by their typing speed. Despite this, he published sparingly. He hated to commit himself to arguments unless he was certain of them. What he did produce however were numerous, and lengthy, drafts of papers and books (at least two of which never saw the light of day) that were widely circulated amongst the philosophical community and even more voluminous comments and responses to other philosophers on how they could improve their arguments. Likening Derek to an iceberg would be mistaken. Up to 10% of an iceberg is above the waterline, whereas I doubt if even 1% of Derek’s work has ever been published. As one of his obituaries noted ‘When Derek Parfit published, it mattered!’

Here is the link, by Simon Beard, and it offers further Parfit anecdotes, via Michael Gibson.  And here is the famed Larissa MacFarquhar profile of Parfit from 2011.

The economic importance of the Middle East?

Or lack thereof?:

Even beyond Iran, the region scarcely registers on multinationals’ profit-and-loss statements. The Middle East and Africa accounted for 2.4% of listed American firms’ revenues in 2019, according to Morgan Stanley, a bank. For European and Japanese companies it was 4.9% and 1.8%, respectively. Middle Easterners still buy comparatively few of the world’s cars (2.3m out of 86m sold globally in 2018). Peddlers of luxury goods like Prada, an Italian fashion house, and L’Oréal, a French beauty giant, book 3% of sales in the Middle East (not counting sheikhs’ shopping trips to Milan or Paris).

The overall regional footprint of Western finance appears equally slight. At the end of 2018 big American banks had $18.5bn-worth of credit and trading activity in the region, equivalent to 0.2% of their assets. This includes JPMorgan Chase’s $5.3bn business in Saudi Arabia and Citigroup’s $9.6bn exposure to the United Arab Emirates (UAE). European banks have, if anything, been retreating. BNP Paribas of France sold its Egyptian business seven years ago and earned a footling €121m ($143m) in the Middle East in 2018. HSBC reports a substantial $58.5bn in Middle Eastern assets, though that is still a rounding error in the British lender’s $2.7trn balance-sheet.

Here is more from The Economist, noting that energy does play a larger role in other economic sectors.  If there is any part of the world that could use more multinational activity…

Africa Luxembourg Ethiopia Congo state capacity facts of the day

Government revenues average about 17% of gdp in sub-Saharan Africa, according to the IMF. Nigeria has more than 300 times as many people as Luxembourg, but collects less tax. If Ethiopia shared out its tax revenues equally, each citizen would get around $80 a year. The government of the Democratic Republic of Congo is so penurious that its annual health spending per person could not buy a copy of this newspaper.

That is from The Economist.