Profile picture
Lyman Stone @lymanstoneky
, 59 tweets, 22 min read Read on Twitter
Today is a big day.

Today is #NBERday

In factm double-#NBER-day, because I wasn't able to get to it last Monday.

Q: Do tax hikes impact economic growth?

A: Yes. They reduce it (in interwar Britain). #NBERday
The method here is fun. They went through every budget in interwar Britain (1918-1939), and categorized each tax by its motivation: ideological, countercyclical, etc. They argue some of these taxes are plausibly exogenous due to arbitrary political motivations. #NBERday
Now, I'm not 100% sure this is true: ideology itself may have some endogeneity with economic conditions. But they do at least show that economic changes do NOT predict their exogenous-labeled taxes, but DO predict other tax changes. #NBERday
Anyways. They find that the implementation of these exogenous taxes had economic effects: 1% of GDP cut in taxes raised GDP by up to 2.5%! That's actually a pretty high multiplier! #NBERday
Now that said, it's still nowhere close to Laffer-equivalence. Revenue effects from induced growth would offset ~25% of lost revenue-ish, at most. Maybe a tiny bit more in the long run. So you can't tax cut your way to budget balance. #NBERday

Q: Does civic leadership have statistically measurable effects?

A: YES. The presence of 48er German refugees dramatically boosted Union recruitment during the civil war. cc @mikeduncan #NBERday
The paper is able to identify 493 specific 48er political refugees who settled in 87 towns around the US. They put together a solid set of controls that satisfies me (i.e. controlling for other German presence). #NBERday
Their finding is remarkable:

Having a 48er in your town basically doubled the Union recruitment rate; i.e. 20 more recruits per hundred resident men. #NBERday
The presence of 48ers also boosted the presence of German-language newspapers and German social organizations. The recruitment effects are specific to German-Americans, with diminution across language lines. Also largest for 48ers who themselves enlisted. #NBERday
But even 48ers who did NOT enlist had a sizable effect! That matters because if it was just enlisted 48ers, it could have been that those enlisted 48ers tended to live in towns that had other active high-quality Union recruiters. #NBERday
The presence of an enlisted 48er in a combat unit even reduced DESERTION rates!

This paper is an incredible testimony to the importance of charismatic, committed leadership. #NBERday
I am impressed by this paper. I'm usually pretty skeptical of IV, and I do worry a bit that 48ers may just coincide with towns that generally appealed to charismatic political Germans (i.e. they may be less randomly distributed than the authors claim).... #NBERday
But consistency of the effect across disparate indicators (recruitment, newspapers, social orgs), including wartime effects that *follow the 48er when they are deployed* (desertion), and the expected influence gradient deterioration across ethnic lines are persuasive. #NBERday

My computer is forcing me to restart.


Okay, quick restart, I'm back.

Q: Can national sports competitions create national togetherness and reduce intranational conflict?

A: Yes! (in Africa)
Basically, this interview looks at surveys taken of Africans JUST BEFORE and JUST AFTER major international soccer matches, focusing on matches expected to be close, and matches which were highly significant (finals). #NBERday
Basically, when the national team wins a high-stakes official game, it causes respondents to express less pride in ethnic identity, and more pride in national identity. Occurs regardless of WHERE the game occured, btw. #NBERday
In other words:

When a country with a weak state has its soccer team do unexpectedly well in a highly salient televised event, it causes people to identify WITH THE COUNTRY.

This is interesting because it's rare to find quasi-causal demonstrations of national identity formation. #NBERday
The effect is also stronger when the defeated team is a traditional rival: so a nearby country or longstanding opponent. #NBERday
The effect of sports is strongest in places with (1) WEAK state presence and (2) HIGH tribal/ethnic division. In other words, sports victories are a key part of nationalism in the places where national identity is currently weakest. That could be good or bad. #NBERday
The authors suggest that this could mean sports fill a key gap in nation-building.

But it could also be that sports-nationalism *is in fact indicative of or perhaps even causally related to* weak state capacity/high division. #NBERday
However, one point in favor of the authors' view that sports nationalism is a driver of positive national feeling, not just an indicator of underdevelopment, is that unexpected Africa cup qualifications *actually reduce political violence* for the ensuing months. #NBERday
Basically, the optimal distribution of sporting outcomes for global economic development is poor countries smearing rich countries, because in areas with high state capacities, losing a sporting event doesn't impact national feeling. #NBERday
BY THE WAY, this reinforces my view that the Moral Purpose Of The Nationals/Cps/Wizards is... to lose. Their place in the universe is ritual defeat and submission to the hinterlands. You should always cheer AGAINST the team from the richer/more powerful place. #NBERday
Unless it's a team from KY in which case it should win. #NBERday
Here's the key chart for the violence response to Africa's cup qualifications: #NBERday
Let's pivot to a new theme: education.

Q: If you reward/punish schools for student attainment, will it cause good teachers to flee bad schools?

A: No. The opposite. The more you punish schools the better their teachers get. #NBERday
This is a study of New York's school accountability system. The theory is that good/bad accountability grades may or may not motivate teachers, but the TOTALLY motivate PRINCIPALS. #NBERday
In other words, when their school is scored badly, principals respond aggressively by trying to improve the school: these improvements are usually amenable to teachers, especially high-quality teachers. #NBERday
But when schools are scored very badly, principals get lazy. They rest on their laurels. They don't implement policies to retain and attract good teachers. So turnover rises and quality falls. #NBERday
In other words... school scorecards have exactly the effect you'd expect. They signal to school administrators when they need to make changes, and administrators do in fact respond. #NBERday
The study design here is a fairly straightforward regression discontinuity design; no major qualms with it from my first reading. #NBERday
But look. There may be simpler ways to improve schools:

Q: Do kids do worse when it's hot in school?

A: Yes. #NBERday
This paper has already made the rounds a fair bit. But basically, hot weather in schools appreciably worsens student performance on tests. Has implications for regional differences in testing, as well as global differences! #NBERday
However, once you add AC, outdoor climate has effectively no impact on test scores. So it's just about temperature in the room where students are learning. #NBERday
The method is pretty decent, the data straightforward, and the mechanism makes sense. I have no quibble with this paper.

Air Conditioning Is Good.
Speaking of education...

Q: Do state merit-based aid grants cause students to substitute away from work towards academics?

A: Yes... but seriously people this effect is really small where are you making some poor economist study this? #NBERday
BASICALLY, the actual question here is, "Does the rise of merit based aid explain the decline in teenage employment?" The answer there is, "a tiny bit, but mostly no." #NBERday
However, honestly, reading their findings, the results seem pretty striking: implementing a selective, merit-based program of average size would reduce age 16-18 school-year LFP by 4 percentage points. That ain't nothin! #NBERday
Accounting for coverage of these programs, seems like state-based merit programs account for ~10% of declining youth LFP! When you add in rise of private-sector merit programs, or university-specific merit programs, it gets substantial! #NBERday
Overall, while the effect size isn't enormous here, if anything this *reinforces* my prior that the college-aid-rat-race is decreasing teen LFP: an hour spent studying earns more in present-value-of-future-aid than an hour of work earns in PV-of-income-plus-interest. #NBERday
So: have "state-provided individual merit-contingent grants" driven declining teen LFP? Well, only a liiiitle bit. But this reinforces the idea that merit aid generally does lower LFP; and private aid is often even more tournament-style and competitive! #NBERday
Speaking of tournament-style returns...

Q: Does winning the lottery make you sad?

A: Not in Sweden! #NBERday
Really not much more to say about this. The lottery curse isn't real guys. #NBERday
Next up:

Q: Why do returns to wealth vary across the wealth distribution; or, alternatively: why do rich families stay rich across generations?

A: Partly, genetics.

ruh roh
Basically, people with lots of wealth, and who get high returns on their wealth, are somewhat genetically different. Across the whole population, their kind of genetic difference predicts risk-takingness. #NBERday
In other words, some people are genetically predisposed to higher-economic-returning behaviors. But crucially, what these genetic traits REALLY predict is people who are very bad at probabilistic thinking. #NBERday
Basically, people with a low genetic score for this indicator which predicts returns to wealth are more likely to give the probability of a given event as "0%", "100%", or "50." High-scorers give numbers like "73%". #NBERday
Basically, this polygenic scoore is predicting people with a specific behavior: weakness at considering probabilistic topics, which then leads to negative experiences, which then leads to strong risk averseness. #NBERday
Now, all too often, people act like, "well if it's genetic there's nothing we can do about it."

Not so!

These polygenic score's impact on economic outcomes is wildly different for people who have defined benefit vs. defined contribution retirement plans. #NBERday
Among the defined benefit group (pensions), the polygenic score was far LESS predictive of retirement savings than among the defined contribution (401ks) group. In other words, the genetic bias can indeed be offset fairly straightforwardly. #NBERday
Now, whether it is *fair* to offset the fact that some people stink at probability by transferring assets, or how much of a transfer is fair, from people who don't stink at probably is a question for another day. But certainly genetic differences are not insurmountable.#NBERday
Last paper:

Q: Do patent examiners show favoritism for patents from companies who eventually employ them?

A: Yes. And those favoritism-impacted patents tend to be lower-quality and get fewer future citations. #NBERday
Basically, patent examiners are subject-matter experts. They could make money in the private sector. Many eventually do exactly that. And it turns out, these revolving-door employees have lower standards for potential future employers (i.e. big firms in their field) #NBERday
Patent examiner standards (measured by ex post quality of patents granted, as measured by citations) get lower when big companies are hiring: i.e. patent examiners give patents more easily at times when they may be applying for outside jobs! #NBERday

This is as bonkers and corrupt as it sounds like. #NBERday
And on that encouraging note, I'm calling it for #NBERday.

As always, I hope you enjoyed, and I encourage you to direct all negative feedback to @Noahpinion .
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Lyman Stone
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member and get exclusive features!

Premium member ($3.00/month or $30.00/year)

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!